i

Analysis and Simulation of Variable

Gain Bandwidth Product Amplifiers
By

M. Maidul Islam
Tasmia Tahmid
Sayeda Salma Nahar

A THESIS
SUBMITTED TO
MILITARY INSTITUTE OF SCIENCE AND TECHNOLOGY

IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR

THE DEGREE OF BACHELOR OF SCIENCE

DEPERTMENT OF ELECTRICAL ELECTRONIC

AND COMMUNICATION ENGINEERING

MIST, DHAKA

DECEMBER, 2006
 ii

Declaration
We declare that this thesis entitled “Analysis and Simulation of
Variable Gain Bandwidth product Amplifiers” is a piece of original research
work. This work has not been presented any where before for the award of
any degree or diploma.

Counter Signed Signature of the Students

………………………….
M. Maidul Islam
………………………. St. No - 200316029
Assistant professor Hamidur Rahman

.......................................
Tasmia Tahmid
St. No – 200316037

.......................................
Sayeda Salma Nahar
St. No – 200316021
 iii

Abstract
This thesis presents the analysis and simulation of variable gain
bandwidth product (GBP) amplifiers. It was previously explored that current
feedback amplifier can provide variable GBP. This does not apply to the all
encountered feedback configurations. In the thesis paper “Novel Design
Approaches for Low–Voltage CMOS Current Feedback and
Transconductance Feedback Amplifier” (-by Md. Ataur Rahman Sarker) we
found some tables (chapter – 2, page - 25) that directed to the way of our
exploration. There we found that there are six possible combinations of the
amplifying elements and the feedback configurations where we can find
constant bandwidth. Then we carried on our research on those configurations
to design practical amplifiers with variable GBP.

In this thesis paper firstly we discussed on the recent developments on

the field of our interest. Then we overviewed the theories concerning
amplifiers that followed by our works where we explored to the way of our
interest.

Then we tried to simulate the circuits which were built by the theories
we derived. After that we discussed the results obtained form the simulation.

Then finally we concluded with the suggestions for the future works
based on our research.
 iv

Acknowledgement

Our sincere and honest gratitude goes to our supervisors’ assistant

professor Hamidur Rahman and assistant professor Md. Ataur Rahman
Sarkar for giving us the opportunity to successfully complete the thesis and
proper guidance throughout our research works.

We are grateful to Wg.cdr Hossam-e-Haider and Ltcol Moin Uddin

for giving us all academic supports during our thesis.

We would like to thank MIST for giving us financial assistance which

was highly essential to pursue our thesis.

Lastly but not the least we would like to thank our parents for their
moral support and encouragement.
 v

Contents
Declaration …………………………………………………………… ii
Abstract …………………………………………………………… iii
Acknowledgement ............................................................................................ iv
Contents ............................................................................................ v

1 Introduction
1.1 Brief history …………………………….. 1
1.2 Recent development in this field .............................................. 4
1.3 Why are we concerned with our work ……………………………. 11
1.4 Outline of the thesis ……………………………. 11

2 Review of previous works

2.1 Introduction ………….…….… 13
2.2 Amplifier and its uses ………..............….. 14
2.3 Amplifier classification …………………... 15
2.4 Concept of VFA and CFA ………………….. 23
2.5 Gain Bandwidth Behavior in Feedback Amplifiers ………………….. 29
2.6 Conclusion ………………….. 34

3 Our works

3.1 Introduction ……………… 35

3.2 Our Proposed topologies to design
constant bandwidth amplifiers ………………... 36
3.3 Topology – 1 ………………… 37
3.4 Topology – 2 ……………… 41
3.5 Topology – 3 ……………... 46
3.6 Topology – 4 ……………... 50
3.7 Bode plots of the gain equations
of the proposed topologies (with the help of MATLAB) ………… 54
 vi

3.8 Conclusion …………………………………………... 69

4 Simulation & Results

4.1 Introduction ………….... 70

4.2 Tools …………... 71
4.3 Simulation of the small signal model ………….. 71
4.4 Simulation of the CMOS design of proposed topology – 1 ……….. 84
4.5 Conclusion ………….. 104

5 Future Works and Conclusion

5.1 Future Works …………………………………. 105

5.2 Conclusion ………………………………… 107

Bibliography ………………………………………………………………. 108

Appendix A .................................................................................................. 110

Appendix B .................................................................................................. 125

Appendix C .................................................................................................. 130

 1

Chapter – 1

Introduction

1.1 Brief history

The process of raising the strength of a weak signal is known
as amplification and the device which accomplishes this job is
called an amplifier. According to the ability of an amplifier to
amplify the voltage level of the input signal or the ability to
amplify the power level of the input signal amplifier can be either
voltage amplifier or power amplifier.

Again among the different analog signal processing devices,

amplifier is most probably the best known device because of its
widespread use and importance. The most common type of the
device is the Voltage Feedback Amplifier (VFA) which represents
the historical dominance of the voltage mode analog design in
electronic industry. Though it is less explored, other types of
feedback amplifiers exist both in industry and academia. Among
these, the Current Feedback Amplifier (CFA) are another
important member in amplifier group and employs the current
mode design approach.
 2

However, the amplifier we use today was not in its present

form in the previous stage. With the invention of the phototube, the
history of signal amplification began in 1875, which was invented
by American G.R.Carey. Day by day, the rapid inventions of X-ray
Tube by Wilhelm Roentgen in 1895, Cathode Ray Oscilloscope by
Karl Braun in 1897, Magnetron Tube by Albert Hull in 1921 and
Klystron Tube by Russell and Sigurd Varian in 1938 became part
of history. With the development of technological advancement,
the vacuum tube has been described as the most important single
piece of equipment introduced into electrical engineering. Its
development has produced a new branch of engineering called
electronics. The applications of vacuum tubes are so varied that
this “miracle tool” has won a place in the industrial and
commercial fields. These tubes have been finding wide
applications in radio, long distance telephones, sound motion
pictures, television, radar, electronic computers and industrial
automation.

Up until 1950s, electronic active device technology was

dominated by the vacuum tube. The major breakthrough in this
trend was experienced in 1947, when the invention of transistors
created a very dramatic change in electronic industry, which was
invented by William B.Shockley, Walter H. Bratt-ain and John
Bardeen of Bell Telephone Laboratories.

In 1960s two stages directly coupled AC Voltage Amplifier

was invented which arrangement is a two stage, directly coupled
AC voltage amplifier.

This has two transistors suppose T1 and T2 in a grounded-

emitter configuration. Isolated Input Data Acquisition Amplifier
also took place in modern technical industry in 1960s.

In 1970s Small Signal Current to Voltage Amplifier,

Adjustable Under frequency over frequency Limiting Circuit, and
 3

Active Bias High Voltage Amplifier became the results of

invention of several other scientists. The small signal current to
voltage amplifier is such a circuit which has application as a
differentiator. In adjustable under frequency over frequency
limiting circuit, the output of the voltage amplifier is, in turn,
applied to the voltage control oscillator over a certain connecting
path. The active bias high voltage amplifier is a cascade amplifier
with an active base biasing circuit which is used to overcome the
classical limitation of using transistors in a high voltage amplifier.
Low Drift AC Coupled Photodiode Amplifier, Voltage Reference
Buffer, Differential Automatic Zero Correction Amplifier,
Automatic Zero Correction Amplifier, Modified ‘H’ Driver Power
Amplifier also became the part of the history.

In 1980s Bias Stabilized High Voltage Amplifier was

invented. In 1990s Dual Amplifier designs behave like a traditional
voltage amplifier and increases the accuracy of current feedback
amplifier. There was also a use of MEDINA FIRM’S AMPLIFIER
IN EXPERIMENTS ABROAD SPACE SHUTTLE; the high-
voltage amplifier was used in several experiments with NASA and
the European Space Agency.

Now in the age of scientific and technological advancement,

that means the time from 2002 to 2006, High-voltage amplifier
driving piezo tubes, and Wideband Amplifier which offers
dynamic range up to 80 dB, are the latest inventions of scientists
FEMTO. GmbH has introduced this new logarithmic wideband
voltage amplifier HLVA-100.This FEMTO device has a wide
dynamic range of up to 80 db.
 4

1.2 Recent Development in this field

An understanding of feedback amplifiers, such as current

feedback amplifiers and voltage feedback amplifiers, are central to
the design of many modern high-speed electronic systems. As
voltage feedback amplifiers, when utilized, have faced a number of
problems as they exhibit a constant gain bandwidth product,
resulting in a bandwidth which reduces in inverse proportion to
gain level. In contrast, the new generation of current feedback
amplifier (CFA) circuits promises improved performance in the
respect of bandwidth remaining constant even at higher levels of
gain. Its topology enables amplifiers to offer unique performance
advantages over traditional voltage feedback amplifiers (VFA).
These advantages are the reason more and more engineers are
designing with CFA designs based on CFA principles have
emerged as the preferred solution for constant bandwidth amplifier
circuits in favor of a current conveyor approaches.

The first current conveyor was developed, unintentionally, in 1966,

when A. S. Sedra was doing research while pursuing his master’s
degree at the University of Toronto. Since then, this device has
been improved and after the first generation, which was totally
based on bipolar technology, it entered into the second generation.
The CCII is a three terminal device represented symbolically in
Figure 2.14. As the name implies, a current conveyor is also called
a current copier. Among the three terminals, X and Z are the
current copying terminals where the current into the terminal X is
copied to the terminal Z.
 5

1.2.1 Proposed CFA Design Topologies

In this section, two different CFA design topologies are
presented which are proposed in the thesis paper “Novel Design
Approaches for Low–Voltage CMOS Current Feedback and
Transconductance Feedback Amplifier” (-by Md. Ataur Rahman Sarker).
These topologies use an opamp to achieve gain by moving the
transimpedance away from the Z node of the CCs. One point to
note here is that, for conventional designs, the high transimpedance
is provided by the resistance and the parasitic capacitance at the Z
node of the CCs. In the proposed designs, since the operational
amplifier accounts for the gain, CFAs designed this way do not
require a high transimpedance. In general, both topologies replace
the buffer with an opamp. In the sections that follow, the design
methods are presented with due explanation.
Topology 1
Firstly, a CFA design is proposed in which a second
generation positive current conveyor (CCII+) is followed by an
opamp. The opamp is operated in open-loop configuration and the
current feedback is provided by means of a single feedback resistor.
The proposed architecture is shown in Figure 3.1. Comparison of
Figure 3.1 with conventional CFA architecture reveals the
following distinct differences: i. the buffer is replaced by the
opamp, ii. The high impedance node is converted to low
impedance node and iii. The opamp in the proposed topology is
operated in open-loop mode. The proposed CFA circuit in
inverting configuration is represented in Figure 3.2. Note that Cz is
the internal capacitance at the Z node of the CCII+. A load
resistance R1 is inserted at the Z node of the CCII+, which results
in a parallel combination with Rz, the resistance at the Z node of
the CCII+, and R1. Since, R1 is very low as compared to Rz (i.e.Rz
_ R1), their parallel combination is effectively R1. Thus the
impedance of the Z node is reduced to _sCz//R1_from _ 1sCz//Rz_.
However, the requirement of high transimpedance is now
compensated by means of the opamp open loop gain, which is
typically very high (>50 dB). The advantage of -
 6

using the opamp is readily understood if we consider the

generation of the high transimpedance in the conventional design.
In order to generate high transimpedance, it needs sufficient
number of cascade transistors between supply rails, which is
problematic for low voltage applications. This is due to the DC
biasing problem of the cascade transistors with low-voltage supply
and eventually some of the transistors might operate out of
saturation. On the other hand, keeping the Z node as a low
impedance point (by using minimum number of cascade transistors
between supply rails) and boosting the low gain by the opamp is an
interesting alternative to compensate the requirement of high
transimpedance.
 7

This is accomplished by driving the CCII+ output current

through the load resistor R1, which is connected to the non
inverting input of the opamp. The voltage across R1 is then
amplified by the opamp’s high open-loop gain. The design is
purposely carried out so that the CCII+ bandwidth is a lot higher
than that of the opamp and hence, the opamp defines the overall
bandwidth of the proposed CFA. Analysis shows that this
arrangement results in a closed-loop bandwidth fixed by the
feedback resistor Rf and load resistance R1. Therefore, the gain in
voltage amplifying configuration can be adjusted without altering
the bandwidth.

Topology 2

The second proposed CFA architecture consists of a second

generation negative current conveyor (CCII-) followed by an
opamp. The design architecture is shown in Figure 3.4. In this case,
the opamp is operated in closed-loop mode and feedback is
arranged only across the opamp, leaving the CCII- out of the
feedback loop. The circuit in Figure 3.5 represents the proposed
CFA in inverting configuration. Note that Cz and Rz are the
 8

internal capacitance and resistance at the Z node of the CCII-,

respectively. In this case, the impedance of the CCII- Z node is
defined by Rz and Cz. Figure 3.6, an extension of the design in
Figure 3.5, incorporates an extra opamp (A2) to reduce the input
resistance at the X terminal of the current conveyor.
 9

Unlike the conventional CFA design, which has a bandwidth

setting feedback resistor around the entire CFA, the proposed
topology uses an opamp instead of buffer and negative feedback is
provided only across the opamp A1.
 10

1.2.2 Proposed TFA Design Topology

The main purpose for initiating research into TFAs is to
realize a low-voltage CMOS version of the device. Keeping this in
mind, a CMOS Wilson TFA was developed. Then the Wilson TFA
is modified by replacing the single-output transconductance block
with a dual-output transconductance block. The proposed design
architecture is presented in Figure 4.1 alongside the Wilson
topology for comparison purpose. A comparison of Figure 4.1 with
Figure 4.2 reveals two clear modifications performed on the
Wilson TFA, namely
• The single-output transconductance cell has been replaced by a
dual-output transconductance cell and
• The unity feedback is provided between one output of the dual-
output transconductance cell and the input of the opamp.

The two outputs of the dual-output transconductance cell are

an exact copy of each other so that they provide exactly the same
 11

signal output. The architecture is translated into circuit level as

shown in Figure 4.3, which is in inverting configuration. It is noted
that the opamp output feeds the positive input of the dual-output
transconductance cell and the opamp is operated in open-loop
configuration. The negative input of the gm cell is always
grounded. Note that co is the parasitic capacitance and ro is the
resistance at the outputs of the dual-output transconductance cell.
By using these topologies the analysis of the behavior of fixed
bandwidth amplifiers was attempted.

1.3 Why are we concerned with our work?

A lot of research has been conducted on feedback amplifiers

having similar gain configuration i.e. for voltage amplifier the gain
configuration used is also a voltage. So we are interested in
analyzing the amplifiers having dissimilar input and gain
configuration. Normally we know that, the gain bandwidth product
is constant. In that case if we want to increase the gain the
bandwidth is reduced, which is not desirable.

We have proposed some amplifier configurations for which

bandwidth is always constant. So that we can utilizes the benefits
of large bandwidth.

1.4 Outline of the thesis

This thesis is organized in chapters and appendices as mentioned
below:
 12

Chapter 1 presents a general introduction of our research

work and the associated research interests.” Brief History”
represents a brief description of how the amplifier we use today
arrives at this present form. From the inventions of phototubes, the
journey of the amplification of signal began and up to the
inventions of high voltage amplifier driving piezo tubes, wideband
amplifier and many other research works of the scientists have
occupied the place in the “Brief History”.” Recent Development in
This Field” is an overview of the ongoing research in our field of
interest. In this topic we mainly focuses On the conventional CFA
design topology and two new topology named “Topology 1” and
“Topology 2” which was developed to overcome the demerits of
using conventional design topology. We also give a description
about the “Wilson TFA design topology” and a “new TFA design
topology” which was proposed to realize a CMOS TFA with
special considerations for low-voltage applications.

Chapter 3 is based on the theoretical analysis of variable gain

bandwidth product (GBP) for four combinations of amplifiers,
where the main amplifying elements and the total feedback
amplifiers are of different types. Here we derived gain equations
for the four combinations of amplifiers. Latter in this chapter we
evaluated derived gain equations with the help of Matlab.

Chapter 4 is a representation of the simulation and results

obtained from the schematic simulations by using the software
Matlab, PSpice and Orcad. The corresponding circuit schematics
are presented and the results generated are discussed.

A formal discourse on our thesis work is concluded in

chapter 5 with presenting the future works to be done.
 13

Chapter - 2

Review of Previous Works

2.1 Introduction

In any research or fieldwork review of basic or fundamental

things in that area is very important and it indicates the starting
point of the chronological development on that field. Confirming
the same aim, this chapter presents the basics of voltage amplifiers
along with feedback amplifiers classification and gain-bandwidth
behavior of different amplifiers.
 14

2.2 Amplifier and its use

Amplifier is a device which magnifies or boosts up an input

signal and gives an amplified output. Fig.2.1 shows a simple block
diagram of an amplifier. Where Si is the input signal and S0 is the
output signal and A is the amplification factor.

The magnitude of S0 depends on A. The input and the output

signals can be voltage or current. Table 2.1 represents the different
combination of Si and S0 and the corresponding nature of the
amplification factor A as determined by various amplifier types
Av,Ai,Az,Ag represents voltage gain, current gain, transimpedance,
transconductance respectively.
 15

Amplifier type Si (Unit) S0 A

(Unit) (Unit)
Voltage Voltage Voltage Av (V/V)
(V) (V)
Current Current Current Ai (A/A)
(A) (A)
Transimpedance Voltage Current Az (V/A)
(V) (A)
Transconductance Current Voltage Ag (A/V)
(A) (V)

Table 2.1: Amplifier types and corresponding signals

The amplifier is used between the microphone and audio

speaker to boost up the weak voice signals. There are actually
amplifiers all around us such as in televisions, computers, portable
compact discs (CD) players and many other devices that may use a
speaker to produce sound.

2.3 Amplifier classification

Amplifiers can be broadly classified into two categories:
1. Voltage amplifier
2. Power amplifier

The voltage amplifier amplifies the voltage level of the input

signal. The voltage amplifiers are designed such that maximum
voltage gain can be achieved. A power amplifier amplifies the
power level of the input signal. A power amplifier can feed a large
power to the load. In order to get a large output power from this
amplifier, the input signal must be large. That is why a power
amplifier is called a large signal amplifier.
 16

A power amplifier cannot amplify the signal input power

directly. It first takes power from the DC power sources connected
to the output circuit and then converts it into useful AC signal
power.
Since our thesis is not related with power amplifiers, only the
voltage amplifiers are discussed here elaborately.

2.3.1 Voltage amplifier configuration

Voltage amplifiers can be configured in two ways:

1. Voltage amplifiers without feedback
2. Voltage amplifiers with feedback

In non feedback system the input does not know what is

happening at the output. If for some reason output changes the
input remains unaffected. It is also known as open loop system.

In feedback amplifier a portion of the output is feedback to

the input through the feedback network. It is also known as closed
loop system.
 17

Theoretical analysis shows that the gains of two cases are

A0 = V0/Vi …………………..………………..(2.1)

Af = V0/Vs =A0/1+βA0 …………………….. (2.2)

Where Vi is the input voltage, Vs is the source voltage, V0 is

the output voltage, A0 is open loop gain, Af is closed loop gain, β is
the feedback factor.
 18

2.3.2 Negative feedback and its advantages

Feedback can be either negative or positive depending on the

relative polarity of the signal being feedback into a circuit. In this
thesis, by feedback we will always mean negative feedback and the
classification henceforth presented is also for negative feedback
amplifier. Negative feedback results in decreased voltage gain. If
feedback signal is opposite polarity to the input signal as shown in
fig 2.3, negative feedback results. While negative feedback results
in reduced overall voltage gain, a number of improvements are
obtained, among them being
1. Higher input impedance
2. Better stabilized voltage gain
3. Improved frequency response
4. Lower output impedance
5. Reduced noise
6. more linear operation

2.3.3 Feedback amplifier classification

According to the feedback configuration, there are four types
of feedback amplifiers, as follows.

2.3.3.1 Voltage amplifier with voltage series

feedback

The main amplifying element in this case is a voltage

amplifier. The feedback connection takes a part of the output
 19

voltage fed back in series with the input signal, resulting in an over
all gain reduction.

If there is no feedback (Vf =0), the voltage gain of amplifier stage

is
Av = V0/Vs = V0/Vi ……………………(2.3)
If a feedback signal is connected in series with the input, then

Vf = βV0 ………………………….(2.4)

V i = V s – Vf ………………….………(2.5)

Since V0 = AvVi = Av (Vs – Vf) = AvVs – AvVf = AvVs – Av (βV0)

Then (1 + βAv) V0 = AvVs …………………………………(2.6)

 20

So the overall voltage gain with feedback is

Avf = V0/Vs = Av/1 + βAv ………………………………………………….(2.7)

Where, Avf is the voltage amplification factor with feedback.

2.3.3.2 Current amplifier with current shunt feedback

The main amplifying element here is a current amplifier. The
feedback network feeds back a part of the output current to the
input.

From fig 2.5 we can write,

If = βI0 …………………………………(2.8)
 21

Ii = Is – If ………………………………………….(2.9)

I0 = Ai Ii ………………………………………...(2.10)

The overall current gain with feedback is

Aif = I0/Is = Ai Ii/1 + βAi ………………………………………(2.11)

2.3.3.3 Transimpedance amplifier with voltage

shunt feedback

In this case the main amplifier is a transimpedance element,

i.e. with current input it produces a voltage output. A portion of
output voltage is feedback to the input in the form of current.
 22

From fig.2.6 we can write

If = βV0 …………………………………….. (2.12)

Ii = Is – If …………………………………… (2.13)

V0 = AzIi ………………………….. (2.14)

Then the transimpedance with feedback is

Azf = V0/Is = Az/1 + βAz ………………. (2.15)

2.3.3.4 Transconductance amplifier with

current series feedback

The main amplifier here is a transconductance element with

voltage feedback.
 23

From fig. we may write,

Vf = βI0 …………………………………….. (2.16)

Vi = V s …………………………... (2.17)

I0 = AgVi …………………………... (2.18)

Transconductance with feedback is

Agf = Ag/(1 + βAg) …………………………….(2.19)

A summary of output, feedback, gain, and feedback factor is

provided for reference in table 2.2 below.

Signal Voltage Current Voltage Current

series series shunt series
Output Voltage Current Voltage Current
Feedback Voltage Current Current Voltage
Gain, A Av Ai Az Ag
Feedback Vf/V0 If/I0 If/V0 Vf/I0
factor, β

Table 2.2: Voltage and current signals in feedback amplifier

2.4 Concept of VFA and CFA

The VFA (voltage feedback amplifier) is that, in which the
feedback signal is proportional to the output voltage irrespective of
the load impedance.
On the other hand, in CFA, the feedback signal is
proportional to the output current irrespective of load impedance.
 24

Designs based on CFA principles have emerged as the preferred

solution for constant bandwidth amplifier circuits in favor of a
current conveyor approach [ref]. A CFA effectively displays the
function of a unity gain voltage buffer between its two input
terminals. One terminal presents high input impedance, while the
second terminal exhibits a very low (ideally zero) input impedance
and constitutes a current feedback terminal. CFAs have
conventionally been designed with a second generation current
conveyor (CCII) followed by a buffer. As the name implies, a
current conveyor is also called a current copier. Among the three
terminals, X and Z are current copying terminals where the current
into the terminal X is copied to the terminal Z. Ideally the current
copying ratio should be unity, but practically it falls below unity as
the frequency is increased. Between terminals Y and X, the device
has a voltage following action. Ideally, whatever voltage appears at
Y is copied to X terminal.

For any analog design it is very important to know for

engineers, which topology has better performance. Many engineers
still refuse to design with CFAs. This is due to a few
misunderstandings which can be easily clarified.
 25

The majority of the opamp circuits are closed loop feedback

system that implements classical control theory analysis. It is
comfortable to design analog circuits using VFA in a closed loop
system. Most circuits commonly built with VFAs can utilize CFAs,
yielding better results at high frequencies. CFAs have sacrificed
the DC precision of VFA in a trade off for increased slew rate and
a bandwidth that is relatively independent of closed loop gain.
Although CFAs do not have the DC precision of their VFA counter
parts, they are good enough to be DC coupled in video applications
without sacrificing too much dynamic range. The days when high
frequency amplifiers had to be AC coupled are gone forever,
because some CFAs are approaching GH gain bandwidth region.
The slew rate of CFAs is not limited by the linear rate of rise that
is seen in VFAs, so it is much faster and leads to faster rise or fall
times and less intermodulation distortion.

The comparison between VFA and CFA is given below.

VFA CFA
1. Lower noise 1. Lower distortion
2. Better DC performance 2. Better linear phase
performance
3. Feedback freedom 3. Feedback restriction
4. Constant GBP 4. BW almost independent of
closed loop gain
5. Relatively lower BW 5. Higher potential BW than
VFA
6. Lower slew rate 6. Higher slew rate
7. Low input offset voltage 7. Nonzero input offset voltage
8. High input impedance 8.Unequal input stage
impedance
9. Matched input bias current 9. Unmatched input bias current
10. Low output impedance 10. High output impedance
 26

The choice of CFA is application specific. In many cases

CFA has a far better result than VFA. Since our thesis is base on
variable GBP amplifiers we have preferred CFA for our further
theoretical analysis.

2.4.1 Gain bandwidth behavior of VFA and CFA

The bandwidth behavior of the two amplifiers can be readily

compared in the following figures. The term ‘loop gain’ in an
electrical circuit is defined as the product of the forward gain and
the feedback factor. For example, in Figure 2.3 loop gain is given
by Ao. Hence, it is clear that loop gain is always associated with a
feedback system.
 27

Figure 2.8 represents the ideal asymptotic characteristic of a

VFA. Here BW1 and BW2 are the bandwidths corresponding to
the demanded gains A1 and A2, respectively. It is clear that when
gain is increased from A1 to A2, bandwidth reduces from BW1 to
BW2. This is because

A1BW1 = A2BW2 …..…………………………….(2.19)

Or
GBP1 = GBP2 ………………………………………………...(2.20)

Where (AiBWi = GBPi) {i = 1, 2}.

This constant gain bandwidth product (GBP) is a result of the

variable loop gain in VFAs. As shown in Figure 2.8, loop gain
(Ao−Ai) {i = 1, 2}, where Ao is the open loop gain, varies with the
 28

demanded gain. For this reason, bandwidth is controlled by the

demanded gain and varies in inverse proportion of the gain.
However, CFAs behave in a totally opposite way as shown in
Figure 2.9.

In this case, the forward path gain automatically increases to

cope with the increasing demanded gain and vice versa. Thus, it
maintains a constant loop gain. The advantage of this phenomenon
is that the gain curve is adjusted automatically In this case, the
forward path gain automatically increases to cope with the
increasing demanded gain and vice versa. Thus, it maintains a
constant loop gain. The advantage of this phenomenon is that the
gain curve is adjusted automatically to accommodate the new level
of demanded gain keeping the closed-loop bandwidth constant. So,
 29

it is clear that this type of amplifier will have variable GBP. As

noticed in Figure 2.9, though the gain is increased from A1 to A2
bandwidth remains constant at BW i.e.

GBP1 ≠ GBP2 ………………………………...(2.21)

And this is due to the constant loop gain TL. In conclusion, the
CFA has a variable GBP and the bandwidth ideally remains
constant irrespective of the demanded closed loop gain.

2.5 Gain Bandwidth Behavior in Feedback

Amplifiers

Recent development in feedback amplifiers shows that

constant bandwidth behavior is the most prevalent case. However,
this does not simply apply to the four most commonly encountered
feedback configurations, where the amplifier and the feedback
arrangement are of the same type, namely, voltage series, current
shunt, voltage shunt and current as shown in Table 2.2 VFAs are
used in general purpose electronic system where constant GBP is
required and the resulting bandwidth is inversely proportion to the
demanded gain level. On the other hand, the current feedback
amplifiers (CFAs) provide better performance with respect to
bandwidth and gain level. The ICs designed using the current
feedback topology realizes constant bandwidth even at higher gain
levels than those in VFAs series. The CFA principle has a greater
benefit that it can be used to achieve constant BW amplifiers.
Constant bandwidth behavior can be achieved by using different
gain configurations as shown in table 2.3, 2.4, 2.5, 2.6. One thing
is to mention here that the type of feedback applied and the type of
 30

amplifying element employed with the feedback network are

independent choices. So we can have a variety of combinations
with respect to the feedback and the amplifying element types as
clarified in Tables 2.3 to 2.6. When amplifier type and the
feedback used are dissimilar the constant bandwidth behavior is
potentially available.

The four basic transfer functions, namely, voltage gain,

current gain, transimpedance and transconductance, along with the
four possible amplifying element types resulting sixteen possible
combinations of the amplifying elements and the feedback
configurations. The type of feedback used will define the stability
of the overall transfer function and the amplifying element type
will govern the input and output impedances as well as the
bandwidth dependency on the demanded gain. For example,
voltage series feedback stabilizes the voltage gain, whereas the
voltage shunt feedback stabilizes the transimpedance, and so on.
The performance of the combined amplifier and feedback network
can be determined by evaluating the closed-loop gain and the loop
gain. If the loop gain varies according to the inverse proportion of
the closed-loop gain, the combined structure will result in constant
GBP. A constant loop gain will produce variable GBP or constant
bandwidth.
 31

The GBP behaviors of all the sixteen possible amplifier and

feedback combinations are summarized in the following Tables.

Amplifying Element Bandwidth

1. Voltage Amplifier GBP constant

2. Current Amplifier BW always constant

3.Transimpedance BW potentially
Amplifier constant
4.Transconductance BW potentially
Amplifier constant

Table 2.3: Gain bandwidth behavior of feedback amplifier in

voltage gain configuration

Amplifying Element Bandwidth

1. Voltage amplifier BW always constant

2.Current Amplifier GBP constant
3. Transimpedance BW potentially
Amplifier constant
4. Transconductance BW potentially
Amplifier constant

Table 2.4: Gain bandwidth behavior of feedback amplifier in

current gain configuration
 32

Amplifying Element Bandwidth

1. Voltage Amplifier BW always constant

2. Current Amplifier BW always constant
3. Transimpedance GBP constant
Amplifier
4. Transconductance BW potentially
Amplifier constant

Table 2.5: Gain bandwidth behavior of feedback amplifier in

transimpedance configuration

Amplifying Element Bandwidth

1. Voltage amplifier BW always constant

2. Current Amplifier BW always constant
3. Transimpedance BW potentially
Amplifier constant
4. Transconductance GBP constant
Amplifier

Table 2.6: Gain bandwidth behavior of feedback amplifier in

transconductance con-figuration

It can be observed from the previous Tables that all the four
cases where the amplifying element and the feedback type are
similar, produces constant GBP. These are the cases most
commonly discussed with feedback amplifiers as summarized in
Table 2.2. For example, a current amplifier designed applying
current shunt feedback around a current amplifying element will
always give constant GBP. Similarly, when voltage,
transimpedance and transconductance amplifying elements are
 33

arranged with voltage series, voltage shunt and current series

feedback networks, respectively, will result in constant GBP. In the
remainder of the cases where dissimilarity is maintained between
the amplifying elements and the feedback types, variable GBP
results, which in turn produces constant bandwidth.

The important feature displayed in Tables 2.3 to 2.6 is that in

all the cases using voltage and current amplifying elements, except
for the cases when the feedbacks are voltage series (Figure 2.4)
and current shunt (Figure 2.5), respectively, display constant loop
gain (TL). Thus, these configurations will provide gain-
independent bandwidth. On the other hand, Tables 2.3 to 2.6 also
clarify that transimpedance and transconductance amplifying
elements produce potentially constant bandwidth, except for the
cases when the feedback arrangements are voltage shunt (Figure
2.6) and current series (Figure 2.7), respectively. The term
‘potentially constant bandwidth’ means that the bandwidth is
constant only under certain specific conditions. This is because
when the main amplifying element is either transimpedance or
transconductance, the loop gain TL depends on one of the elements
that define the closed loop gain. Hence, constant bandwidth
behavior can be experienced only when one of the gain defining
elements is kept constant. CFAs and TFAs fall under these
conditions.
This topic is highly important because the main focus of
our thesis is to analyze and simulate constant BW amplifiers.
 34

2.6 Conclusion

The review of previous works helps us to recapitulate the

basic ideas of our thesis related topics. The key point of this
chapter is that gain-independent BW can be achieved from many
different alternatives where the main amplifying element and the
feedback connections are of different type. For same type of
amplifier and the feedback arrangement GBP is always constant.
The main focus of our thesis is the simulation and theoretical
analysis of sixteen possible combinations [table] of the amplifying
elements and the feedback configuration.
 35

Chapter – 3

Our works

3.1 Introduction
In the previous chapter we saw that current feedback
amplifier has a variable gain bandwidth product and ideally the
bandwidth remains constant irrespective of the closed loop gain it
is providing. The ICs which are designed employing current
feedback topology can provide constant bandwidth even at higher
gain levels than those in the voltage feedback amplifier. But this
phenomenon does not apply to the four most commonly used
feedback configurations, where the amplifier and the feedback
arrangement are of the same type which we observed in the
previous chapter. In the previous chapter we saw that the constant
bandwidth behavior is potentially obtainable whenever the
amplifier type and the feedback used are not similar. There we saw
in the tables under the topic “Gain bandwidth behavior in feedback
amplifier” that in every case (except for the cases where the
feedback is voltage series in voltage amplifier and current shunt in
current amplifier) using voltage and current amplifying elements
provide constant loop gain. This is how these configurations will
provide gain independent band width without having any effect of
the gain defining elements. Constant bandwidth behavior can be
experienced only when one of the gain defining elements is kept
constant.
 36

3.2 Our Proposed topologies to design constant

bandwidth amplifiers

In the tables under the topic “Gain bandwidth behavior in

feedback amplifier” of the previous chapter we can note that there
are six possible amplifiers and feedback combinations where
constant bandwidth is obtained. Our proposed topologies are based
on those combinations. First we worked at the block diagram level
to develop our topology. Then we went to the small signal analysis
from the block diagram level. From the small signal analysis we
deduced the gain equation and the 3dB bandwidth equation for our
proposed topologies. But for the last two combinations which were
– voltage amplifier and current amplifier with such a feed back that
the gain bandwidth behavior of the feedback amplifier is in
transconductance configuration; we obtained the gain equation and
the 3dB bandwidth equation in cumbersome and complicated. We
judged that derived equations of the last two combinations will fail
to provide constant band width behavior. So we did not propose
any topology based on the last two combinations. We proposed
four topologies in total. These four topologies are described one
after another in the following.
 37

3.3 Topology – 1

In the first topology we used a current amplifier as the

amplifying element with such a feedback structure that total
combination works with the gain band width behavior in voltage
gain configuration. In this topology we used a positive second
generation current conveyor (CCII+), current amplifier,
transimpedance block and a feedback network. A current
proportional to input voltage signal appears to the input of the
CCII+, which copies the input current to its output. Output current
of the CCII+ will be amplified by the current amplifier. The output
current of the current amplifier is then used to produce a voltage
proportional to the current itself. This is performed by a
transimpedance block. Finally the feedback network provides
current feedback proportional to the output voltage to the input
terminal of the CCII+. The block diagram is shown below.

Fig 3.1: Block Diagram of proposed topology – 1 using amplifying

element: current amplifier and total feedback amplifier is in
voltage gain configuration to obtain Variable GBP(constant
bandwidth).
 38

3.3.1 Theoretical Analysis

Now we will discuss about the small signal model of our

proposed topology – 1 in inverting configuration. Following figure
shows the small signal model of our proposed topology – 1 in
inverting configuration.

Fig 3.2: Proposed topology – 1 small signal model in inverting

configuration.

Voltage at node –X of CCII+ is 0, because the input of the

buffer at the input side of CCII+ is zero and it copies its input
voltage at node –X. We assumed that the input resistance rx of the
CCII+ is very small and close to zero, so we neglected its effect.
The internal capacitance at the Z node of the CCII+ is Cz and the
resistance at the Z node is Rz. The value of Cz is in pico farad
range and the value of Rz is in several kilo ohm range.
 39

Here we assumed ideal cases for the current amplifier that is -

we assumed input resistance (RiA) of the current amplifier is small
and output resistance (Rio) tends to infinity. So we considered that
a negligible amount of current passes through output resistance
(Rio).

The output resistance Ro works as the transimpedance block.

One portion of the output current passes through it, as a result we
obtain a output voltage Vo proportional to the current io which
passes through Ro. We can use a buffer at the output or for a fixed
load we can set Ro = Rload (Load resistance).

The current feedback is provided through the resistance Rf.

The feedback current if is proportional to output voltage Vo.

Feedback current, if = Vo ;
Rf
Input current to current conveyor, iin = Vin ;
Ri
The voltage gain of the total combination of the proposed topology
– 1 is given as follows -

Gv = Vo = − AiωtRoRf ………………………………..(3.1)
Vin Ri{(S + ωt )( Rf + Ro) + AiωtRo}

The derivation of the above equation is shown in the appendix A.1.

Equation 2.1 reveals that the 3dB bandwidth of our proposed

topology-1 is given by [appendix A.1] –

f = ft{1+ Ai ⋅ Ro } ...................................................................(3.2)
Rf + Ro
 40

From the equation 2.1 (the gain equation of the proposed

topology -1) we can note that feedback resistance Rf, output
resistance Ro and the input resistance Ri are the gain defining
elements which we can vary. But from the equation 2.2 (3dB
bandwidth equation) it is evident that feedback resistance Rf and
output resistance Ro also determines the 3dB bandwidth of our
proposed topology – 1. Input resistance Ri is the only gain defining
element which does not effect 3dB bandwidth. We can conclude
that here we have obtained variable GBP (constant bandwidth), if
the voltage gain is varied by the input resistance Ri.

Small signal model of our proposed topology – 1 in non-inverting

configuration is shown below.

Fig3.3: Proposed topology – 1 small signal model in non-inverting

configuration.

Voltage gain equation for the non-inverting configuration will be –

Gv = Vo = AiωtRoRf …………………………………(3.3)
Vin Ri{(S + ωt )( Rf + Ro) + AiωtRo}
 41

The 3dB bandwidth for the non-inverting configuration will be

same as the inverting configuration, that is-

f = ft{1+ Ai ⋅ Ro }
Rf + Ro

3.4 Topology – 2

In the second topology we used a voltage amplifier as the

amplifying element with such a feedback structure that total
combination works with the gain band width behavior in current
gain configuration. In this topology we used a positive second
generation current conveyor (CCII+), voltage amplifier,
transconductance block and a feedback network. Input current
signal is applied to the input of the CCII+, which copies the input
current to its output. Output current of the CCII+ will be used to
produce a voltage proportional to the current itself. This is done by
the impedance at the node Z of the CCII+. As ideally input
impedance of the voltage amplifier is infinity, the output current of
the CCII+ will flow through the impedance of the Z node. This
will produce the required input voltage of the voltage amplifier.
This voltage is than amplified by a voltage amplifier. The output
voltage of the voltage amplifier is then used to produce a current
proportional to the voltage itself. This is performed by a
transconductance block. Finally the feedback network provides
current feedback proportional to the output current to the input
 42

terminal of the CCII+. Following is the block diagram of the

proposed topology - 2.

Fig3.4: Block Diagram of proposed topology – 2 using amplifying

element: voltage amplifier and total feedback amplifier is in
current gain configuration to obtain Variable GBP(constant
bandwidth).

3.4.1 Theoretical Analysis

Now we will discuss about the small signal model of our

proposed topology – 2 in inverting configuration. Following figure
shows the small signal model of our proposed topology – 2 in
inverting configuration.

Fig 3.5: Proposed topology – 2 small signal model in inverting

configuration.
 43

Voltage at node –X of CCII+ is 0, because the input of the

buffer at the input side of CCII+ is zero and it copies its input
voltage at node –X. We assumed that the input resistance rx of the
CCII+ is very small and close to zero, so we neglected its effect.
The internal capacitance at the Z node of the CCII+ is Cz and the
resistance at the Z node is Rz. The value of Cz is in pico farad
range and the value of Rz is in several kilo ohm range.

Here we assumed ideal cases for the voltage amplifier that is

- we assumed input resistance (RiA) of the voltage amplifier tends
to infinity and output resistance (RoA) tends to zero. So we
considered that a negligible amount of current passes through input
resistance (RiA) and the voltage drop across the output resistance
(RoA) is also negligible.

We assumed ideal cases for the transconductance block that

is - we assumed input resistance (Rig) and output resistance (Rog) of
the transconductance block tends to infinity. So we considered that
a negligible amount of current passes through both the input
resistance (Rig) and the output resistance (Rog).

The current feedback is provided through the resistance Rf.

The feedback current if is proportional to output current io.

Input current to current conveyor is iin. Feedback current, if = Vo .

Rf
 44

The current gain of the total combination of the proposed topology

– 2 is given as follows –

Gi = io
iin
=− RfRzAoωtgm
RoRzAoωtgm + ωt ( Rf + Ro) + S (1+ ωtCzRz)( Rf + Ro) + S 2CzRz( Rf + Ro)
………………………………..(3.4)

The derivation of the above equation is shown in the appendix A.2.

Equation 2.4 reveals that the 3dB bandwidth of our proposed

topology-2 is given by [appendix A.2] –

−(1+ 2π ftCzRz) ± (1+ 2π ftCzRz)2 + 8π ftCzRz ×{1+ RoRzAogm}

( Rf + Ro)
⇒f=
4π CzRz

........................................................(3.5)

From the equation 2.4(the gain equation of the proposed

topology -2) we can note that feedback resistance Rf and out
resistance Ro are the gain defining elements which we can vary.
But from the equation 2.5 (3dB bandwidth equation) it is evident
that feedback resistance Rf and output resistance Ro also
determines the 3dB bandwidth of our proposed topology – 2. From
the 3dB bandwidth equation we can see that we have a chance to
have negligible effect on the 3dB bandwidth for variation of Rf , if
Ro is very much greater than Rf . In this case we can neglect the
RoRzAogm
term 8π ftCzRz ×{1+ } . With the help of MATLAB we will
( Rf + Ro)
 45

observe the effect of variation of various external resistances on

3dB bandwidth by the equation we have obtained.

Small signal model of our proposed topology – 2 in non-inverting

configuration is shown below.

Fig3.6: Proposed topology – 2 small signal model in non-inverting

configuration.

Current gain equation for the non-inverting configuration will be –

Gi = RfRzAoωtgm
RoRzAoωtgm + ωt ( Rf + Ro) + S (1+ ωtCzRz)( Rf + Ro) + S 2CzRz( Rf + Ro)
…………………………………(3.6)

The 3dB bandwidth for the non-inverting configuration will be

same as the inverting configuration, that is-

−(1+ 2π ftCzRz) ± (1+ 2π ftCzRz)2 + 8π ftCzRz ×{1+ RoRzAogm}

( Rf + Ro)
⇒f=
4π CzRz
 46

3.5 Topology – 3

In the third topology we used a voltage amplifier as the

amplifying element with such a feedback structure that total
combination works with the gain band width behavior in
transimpedance gain configuration. In this topology we used a
positive second generation current conveyor (CCII+), voltage
amplifier and a feedback network. Input current signal is applied to
the input of the CCII+, which copies the input current to its output.
Output current of the CCII+ will be used to produce a voltage
proportional to the current itself. This is done by the impedance at
the node Z of the CCII+. As ideally input impedance of the voltage
amplifier is infinity, the output current of the CCII+ will flow
through the impedance of the Z node. This will produce the
required input voltage of the voltage amplifier. This voltage is than
amplified by a voltage amplifier. Finally the feedback network
provides current feedback proportional to the output voltage to the
input terminal of the CCII+. The block diagram is shown below.

Fig 3.7: Block Diagram of proposed topology – 3 using amplifying

element: voltage amplifier and total feedback amplifier is in
transimpedance configuration to obtain Variable GBP(constant
bandwidth).
 47

3.5.1 Theoretical Analysis

Now we will discuss about the small signal model of our

proposed topology – 3 in inverting configuration. Following figure
shows the small signal model of our proposed topology – 3 in
inverting configuration.

Fig3.8: Proposed topology – 3 small signal model in inverting

configuration.

Voltage at node –X of CCII+ is 0, because the input of the

buffer at the input side of CCII+ is zero and it copies its input
voltage at node –X. We assumed that the input resistance rx of the
CCII+ is very small and close to zero, so we neglected its effect.
The internal capacitance at the Z node of the CCII+ is Cz and the
resistance at the Z node is Rz. The value of Cz is in pico farad
range and the value of Rz is in several kilo ohm range.
 48

Here we assumed ideal cases for the voltage amplifier that is

- we assumed input resistance (RiA) of the voltage amplifier tends
to infinity and output resistance (RoA) tends to zero. So we
considered that a negligible amount of current passes through input
resistance (RiA) and the voltage drop across the output resistance
(RoA) is also negligible.

The current feedback is provided through the resistance Rf.

The feedback current if is proportional to output voltage Vo.
Input current to current conveyor is iin. Feedback current, if = Vo ;
Rf
The gain of the total combination of the proposed topology – 3 is
given as follows -

⇒ Rm = Vo = − RfRzAoωt
iin Rf ωt + RzAoωt + S 2 RfCzRz + SRf (1+ ωtCzRz)
………………………..(3.7)

The derivation of the above equation is shown in the appendix A.3.

Equation 2.7 reveals that the 3dB bandwidth of our proposed

topology-3 is given by [appendix A.3] –

− Rf (1+ 2π ftCzRz) ± Rf 2 (1+ 2π ftCzRz)2 + 8π ftCzRzRf ( Rf + RzAo)

⇒f=
4π CzRzRf
......................(3.8)

From the equation 2.7(the gain equation of the proposed

topology -3) we can note that feedback resistance Rf is the only
gain defining elements which we can vary. But from the equation
2.8 (3dB bandwidth equation) it is evident that feedback resistance
Rf also determines the 3dB bandwidth of our proposed topology –
3. From the 3dB bandwidth equation we can see that we have a
 49

chance to have negligible effect on the 3dB bandwidth for

variation of Rf, if 8π ftCzRzRf ( Rf + RzAo) can be neglected from
3dB bandwidth equation. With the help of MATLAB we will
observe the effect of variation of the external resistance Rf on 3dB
bandwidth by the equation we have obtained.

Small signal model of our proposed topology – 3 in non-inverting

configuration is shown below.

Fig3.9: Proposed topology – 3 small signal model in non-inverting

configuration.

Transimpedance gain equation for the non-inverting configuration

will be –

Rm = RfRzAoωt …………… ………(3.9)

Rf ωt + RzAoωt + S 2 RfCzRz + SRf (1+ ωtCzRz)

The 3dB bandwidth for the non-inverting configuration will be

same as the inverting configuration, that is-
− Rf (1+ 2π ftCzRz ) ± Rf 2 (1+ 2π ftCzRz)2 + 8π ftCzRzRf ( Rf + RzAo)
f=
4π CzRzRf
 50

3.6 Topology – 4

In the fourth topology we used a current amplifier as the

amplifying element with such a feedback structure that total
combination works with the gain band width behavior in
transimpedance gain configuration. In this topology we used a
positive second generation current conveyor (CCII+), current
amplifier, transimpedance block and a feedback network. Input
current signal is applied to the input of the CCII+, which copies the
input current to its output. Output current of the CCII+ will be
amplified by the current amplifier. The output current of the
current amplifier is then used to produce a voltage proportional to
the current itself. This is performed by a transimpedance block.
Finally the feedback network provides current feedback
proportional to the output voltage to the input terminal of the
CCII+. The block diagram is shown below.

Fig 3.10 Block Diagram of proposed topology –4 using amplifying

element: current amplifier and total feedback amplifier is in
transimpedance configuration to obtain Variable GBP(constant
bandwidth).
 51

3.6.1 Theoretical Analysis

Now we will discuss about the small signal model of our

proposed topology – 4 in inverting configuration. Following figure
shows the small signal model of our proposed topology – 4 in
inverting configuration.

Fig 3.11: Proposed topology – 4 small signal model in inverting

configuration.

Voltage at node –X of CCII+ is 0, because the input of the

buffer at the input side of CCII+ is zero and it copies its input
voltage at node –X. We assumed that the input resistance rx of the
CCII+ is very small and close to zero, so we neglected its effect.
The internal capacitance at the Z node of the CCII+ is Cz and the
resistance at the Z node is Rz. The value of Cz is in pico farad
range and the value of Rz is in several kilo ohm range.

Here we assumed ideal cases for the current amplifier that is -

we assumed input resistance (RiA) of the current amplifier is small
and output resistance (Rio) tends to infinity. So we considered that
a negligible amount of current passes through output resistance
(Rio).
 52

We assumed ideal cases for the transimpedance block that is

- we assumed input resistance (RiA’) and output resistance (RoA’) of
the transimpedance block tends to Zero. So we considered that
voltage drop across the output resistance (RoA’) is negligible. Let
the gain of the transimpedance block is Rm(S).

The current feedback is provided through the resistance Rf.

The feedback current if is proportional to output voltage Vo.

Input current to current conveyor is iin. Feedback current, if = Vo .

Rf
The transimpedance gain of the total combination of the proposed
topology – 4 is given as follows -

Rm = Vo = − RfRmAiωt ………..(3.10)
iin Rf ωt + RmAiωt + S 2CzRiARf + SRf (1+ CzRiAωt )

The derivation of the above equation is shown in the appendix A.4.

Equation 2.10 reveals that the 3dB bandwidth of our proposed

topology-4 is given by [appendix A.4] –

− Rf (1+ 2π ftCzRiA) ± Rf 2 (1+ 2π ftCzRiA)2 + 8π ftCzRiARf ( Rf + RmAi)

f=
4π CzRiARf
.....................(3.11)

From the equation 2.10(the gain equation of the proposed topology

-4) we can note that feedback resistance Rf is the only gain
defining elements which we can vary. But from the equation 2.11
(3dB bandwidth equation) it is evident that feedback resistance Rf
also determines the 3dB bandwidth of our proposed topology – 4.
From the 3dB bandwidth equation we can see that we have a
chance to have negligible effect on the 3dB bandwidth for
 53

variation of Rf, if 8π ftCzRiARf ( Rf + RmAi) can be neglected from

3dB bandwidth equation. With the help of MATLAB we will
observe the effect of variation of the external resistance Rf on 3dB
bandwidth by the equation we have obtained.

Small signal model of our proposed topology – 4 in non-inverting

configuration is shown below.

Fig3.12: Proposed topology – 4 small signal model in non-

inverting configuration.

Transimpedance gain equation for the non-inverting configuration

will be –

Rm = RfRmAiωt …………………(3.12)
Rf ωt + RmAiωt + S CzRiARf + SRf (1+ CzRiAωt )
2

The 3dB bandwidth for the non-inverting configuration will be

same as the inverting configuration, that is-

− Rf (1+ 2π ftCzRiA) ± Rf 2 (1+ 2π ftCzRiA)2 + 8π ftCzRiARf ( Rf + RmAi)

⇒f=
4π CzRiARf
 54

3.7 Bode plots of the gain equations of the

proposed topologies (with the help of MATLAB)

3.7.1 Bode plot of the topology – 1

Gain equation or the transfer function of the proposed

topology -1 is –

Gv = Vo = − AiωtRoRf
Vin Ri{(S + ωt )( Rf + Ro) + AiωtRo}

Following are the different parameters of our proposed

topology – 1 with their values which are used as the initial input of
the MATLAB program to obtain bode plot.

Gain of the current amplifier, Ai =10

Bandwidth of the current amplifier, ωt=1000 Hz
Output resistance, Ro=100 Ω
Feedback resistance Rf =1000 Ω
Input resistance Ri =100 Ω
 55

With the values of the parameters mentioned above we

obtained the bode plot shown below.

Fig 3.13: Bode plot for the values mentioned initially.

Now we will vary the input resistance Ri. The bode plot
below shows the variation of Ri with the other parameters remain
unchanged as defined initially. Used values of the input resistance
Ri in the bode plot are 10Ω, 100Ω, 1kΩ, 10kΩ.
 56

Fig 3.14: Bode plots showing the effect of variation of Ri.

Now we will vary the feedback resistance Rf. The bode plot
below shows the variation of Rf with the other parameters remain
unchanged as defined initially. Used values of the feedback
resistance Rf in the bode plot are 10Ω, 100Ω, 1kΩ, 10kΩ.
 57

Fig 3.15: Bode plots showing the effect of variation of Rf.

Now we will vary the output resistance Ro. The bode plot
below shows the variation of Ro with the other parameters remain
unchanged as defined initially. Used values of the output resistance
Ro in the bode plot are 10Ω, 100Ω, 1kΩ, 10kΩ.
 58

Fig 3.16: Bode plots showing the effect of variation of Ro.

Our observation on the above bode plots found that for the
variation of input resistance Ri there is no variation in the 3dB
bandwidth. But constant bandwidth is not available for the
variation the other external resistances. This supports our previous
comments on the gain bandwidth behavior of the toplogy-1, which
we did by observing the gain equation.
The MATLAB code for the above bode plots is provided in
the appendix – B.1.
 59

3.7.2 Bode plot of the topology – 2

Gain equation or the transfer function of the proposed

topology -2 is –

Gi = − RfRzAoωtgm
RoRzAoωtgm + ωt ( Rf + Ro) + S (1+ ωtCzRz)( Rf + Ro) + S 2CzRz( Rf + Ro)

Following are the different parameters of our proposed

topology – 2 with their values which are used as the initial input of
the MATLAB program to obtain bode plot.

Gain of the voltage amplifier, Ao =10

Gain of the transconductance block, Gm =5
Bandwidth of the voltage amplifier, ωt=1000 Hz
Output resistance, Ro=100 Ω
Feedback resistance, Rf =1000 Ω
Resistance at node Z, Rz =10KΩ
Capacitance at node Z, Cz =1pF
 60

With the values of the parameters mentioned above we

obtained the bode plot shown below.

Fig 3.17: Bode plot for the values mentioned initially.

Now we will vary the feedback resistance Rf. The bode plot
below shows the variation of Rf with the other parameters remain
unchanged as defined initially. Used values of the feedback
resistance Rf in the bode plot are 10Ω, 100Ω, 1kΩ, 10kΩ.
 61

Fig 3.18: Bode plots showing the effect of variation of Rf.

Now we will vary the output resistance Ro. The bode plot
below shows the variation of Ro with the other parameters remain
unchanged as defined initially. Used values of the output resistance
Ro in the bode plot are 10Ω, 100Ω, 1kΩ, 10kΩ.
 62

Fig 3.19: Bode plots showing the effect of variation of Ro.

Our observation of the above bode plots found that constant

bandwidth is not available for the variation of any external
resistances. This supports our previous comments on the gain
bandwidth behavior of the toplogy-2, which we did by observing
the gain equation.
The MATLAB code for the above bode plots is provided in
the appendix – B.2.
 63

3.7.3 Bode plot of the topology – 3

Gain equation or the transfer function of the proposed

topology -3 is –

Rm = − RfRzAoωt
Rf ωt + RzAoωt + S 2 RfCzRz + SRf (1+ ωtCzRz)

Following are the different parameters of our proposed

topology – 3 with their values which are used as the initial input of
the MATLAB program to obtain bode plot.

Gain of the voltage amplifier, Ao =10

Bandwidth of the voltage amplifier, ωt=1000 Hz
Feedback resistance, Rf =1000 Ω
Resistance at node Z, Rz =1KΩ
Capacitance at node Z, Cz =1pF
 64

With the values of the parameters mentioned above we

obtained the bode plot shown below.

Fig 3.20: Bode plot for the values mentioned initially.

Now we will vary the feedback resistance Rf. The bode plot
below shows the variation of Rf with the other parameters remain
unchanged as defined initially. Used values of the feedback
resistance Rf in the bode plot are 10Ω, 100Ω, 1kΩ, 10kΩ.
 65

Fig 3.21: Bode plots showing the effect of variation of Rf.

Our observation of the above bode plots found that constant

bandwidth is not available for the variation of any external
resistances. This supports our previous comments on the gain
bandwidth behavior of the toplogy-3, which we did by observing
the gain equation.
The MATLAB code for the above bode plots is provided in
the appendix – B.3.
 66

3.7.4 Bode plot of the topology – 4

Gain equation or the transfer function of the proposed

topology -4 is –

Rm = − RfRmAiωt
Rf ωt + RmAiωt + S CzRiARf + SRf (1+ CzRiAωt )
2

Following are the different parameters of our proposed

topology – 4 with their values which are used as the initial input of
the MATLAB program to obtain bode plot.

Gain of the current amplifier, Ai =10

Gain of the transimpedance block, Gm =5
Bandwidth of the current amplifier, ωt=1000 Hz
Feedback resistance, Rf =1000 Ω
Input resistance of the current amplifier, RiA =10Ω
Capacitance at node Z, Cz =1pF
 67

With the values of the parameters mentioned above we

obtained the bode plot shown below.

Fig 3.22: Bode plot for the values mentioned initially.

Now we will vary the feedback resistance Rf. The bode plot
below shows the variation of Rf with the other parameters remain
unchanged as defined initially. Used values of the feedback
resistance Rf in the bode plot are 10Ω, 100Ω, 1kΩ, 10kΩ.
 68

Fig 3.23: Bode plots showing the effect of variation of Rf.

Our observation of the above bode plots found that constant

bandwidth is not available for the variation of any external
resistances. This supports our previous comments on the gain
bandwidth behavior of the toplogy-4, which we did by observing
the gain equation.
The MATLAB code for the above bode plots is provided in
the appendix – B.4.
 69

3.8 Conclusion

From the bode plots of the preceding sections, we can note

that only for topology – 1 we can find a gain defining element
which has no effect on the 3dB bandwidth. The gain defining
element is the input resistance Ri. For the other topologies we
could not obtain any gain defining element which has no effect on
the 3dB bandwidth. But after numerous attempts with different sets
of values of gain defining element to obtain minimum effect on the
3dB bandwidth may result to a considerable point. But still that
may be quite large in comparison with proposed topology – 1.
 70

Chapter – 4

Simulation & Results

4.1 Introduction

In the preceding chapter we saw that our proposed topology –

1 provides the best gain equation to obtain constant bandwidth. It
has only two main circuit components – one is positive second
generation current conveyor (CCII+) and other one is current
amplifier. The transimpedance component can be implemented
with only one resistance. So it will require lesser MOS devices for
CMOS design. Proposed topology – 2 has also two main circuit
components, but its gain equation is not as perfect as the proposed
topology – 1 for obtaining variable GBP. So we thought that it will
be wiser to concentrate on the topology – 1 to obtain any
successful result within a limited time. Accordingly we further
analyzed and simulated only the topology – 1.
 71

4.2 Tools
We used PSpice student version and orcad 9.1 (Evolution
Version) to simulate the circuits under our research. But PSpice
student version can not simulate the schematics having larger
circuits. But it can generate schematics netlist of larger circuits. All
the CMOS designs of our concern for this thesis fall in larger
circuit category. So, we used PSpice student version to obtain
schematic netlist. Because writing schematic netlist is cumbersome
to us for the larger circuits. Ultimately the obtained netlists are
simulated in the orcad 9.1. Actually we tried to use the resources
available to us.

4.3 Simulation of the small signal model

We starred our simulation purpose from the small signal

model. First we simulated the small signal model of the proposed
topology – 1 in inverting configuration which is shown in the
figure below.

Fig 4.1: Small signal model of the proposed topology – 1 in

inverting configuration.
 72

We could not find any buffer in the in the PSpice parts

library. So, we used an ideal opamp in voltage follower mode to
serve the purpose of a buffer. The PSpice Schematic is shown
bellow.

Fig 4.2: Small signal model of the proposed topology – 1 in

inverting configuration. Here the purpose of the buffer is done by
an ideal opamp.

The outputs of the transient analysis of the small signal model of

the proposed topology – 1 in inverting configuration are shown in
the following.

Following figure shows the current flowing through the input

resistance and output resistance. I(R1) is the current flowing
through the input resistance and I(R8) is the current flowing
through the output resistance.
 73

Fig 4.3: I(R1) is the current flowing through the input resistance
and I(R8) is the current flowing through the output resistance.

Following figure shows the input voltage and output voltage

of the topology – 1 for inverting configuration. V(vin) is the input
voltage and V(vout) is the output voltage.

Fig 4.4: V(vin) is the input voltage and V(vout) is the output
voltage.
 74

Following figure shows small signal model of the proposed

topology – 1 in non-inverting configuration.

Fig 4.5: Small signal model of the proposed topology – 1 in non-

inverting configuration.

As we could not find any buffer in the in the PSpice parts

library, we used the following circuit for the small signal analysis
of the proposed topology – 1 for non-inverting configuration.

Fig 4.6: Small signal model of the proposed topology – 1 in non-

inverting configuration. Here the purpose of the buffer is done by
an ideal opamp.
 75

The outputs of the transient analysis of the small signal model of

the proposed topology – 1 in non-inverting configuration are
shown in the following.

Following figure shows the current flowing through the input

resistance and output resistance. I(R7) is the current flowing
through the input resistance and I(R6) is the current flowing
through the output resistance.

Fig 4.7: I(R7) is the current flowing through the input resistance
and I(R6) is the current flowing through the output resistance.

Following figure shows the input voltage and output voltage

of the topology – 1 for non-inverting configuration. V(vin) is the
input voltage and V(vout) is the output voltage.
 76

Fig 4.8: V(vin) is the input voltage and V(vout) is the output
voltage.

From the above figures we can see that we have certainly

obtained a considerable gain for both for inverting and non-
inverting configurations of our proposed topology – 1.

Now we will concentrate on the AC analyses. The following

AC analyses shown are for inverting configuration. Her we will
not show AC analysis for non-inverting configuration because both
will result the same.

Following are the different parameters of our proposed

topology – 1 with there values which are used as the initial values
for the AC analysis.

Gain of the current amplifier, Ai =10

Bandwidth of the current amplifier, ωt=1000 Hz
Output resistance, Ro=100 Ω
Feedback resistance Rf =1000 Ω
Input resistance Ri =100 Ω
 77

With the values of the parameters mentioned above we

obtained the frequency response curve shown below.

Fig 4.9: Frequency response curve for the values mentioned

initially.

Red dot on the frequency response curves indicates the 3dB

point.

Now we will vary the input resistance Ri. Following

frequency response curves show the variation of Ri with the other
parameters remain unchanged as defined initially. Used values of
the input resistance Ri in the bode plot are 10Ω, 100Ω, 1kΩ, 10kΩ.
 78

Fig 4.10-a: Frequency response curve for topology – 1 when Ri =

10Ω.

Fig 4.10-b: Frequency response curve for topology – 1 when Ri =

100Ω.
 79

Fig4.10-c: Frequency response curve for topology – 1 when Ri =

1kΩ.

Fig 4.10-d: Frequency response curve for topology – 1 when Ri =

10k Ω.All the above bode plots showed the effect of variation of Ri.
 80

Now we will vary the feedback resistance Rf. The bode plot
below shows the variation of Rf with the other parameters remain
unchanged as defined initially. Used values of the feedback
resistance Rf in the bode plot are 10Ω, 100Ω, 1kΩ, 10kΩ.

Fig 4.11-a: Frequency response curve for topology – 1 when Rf =

100Ω.

Fig 4.11-b: Frequency response curve for topology – 1 when Rf =

1kΩ.
 81

Fig 4.11- c: Frequency response curve for topology – 1 when Rf =

10kΩ . All the above bode plots showed the effect of variation of Rf.

Now we will vary the output resistance Ro. The bode plot
below shows the variation of Ro with the other parameters remain
unchanged as defined initially. Used values of the output resistance
Ro in the bode plot are 10Ω, 100Ω, 1kΩ, 10kΩ.

Fig 4.12-a: Frequency response curve for topology – 1 when Ro =

10Ω.
 82

Fig 4.12-b: Frequency response curve for topology – 1 when Ro =

100Ω.

Fig 4.12-c: Frequency response curve for topology – 1 when Ro =

1kΩ.
 83

Fig 4.12-d: Frequency response curve for topology – 1 when Ro =

10kΩ . All the above bode plots showed the effect of variation of Ro.

Our observation of the above frequency response curves

found that for the variation of input resistance Ri there is a very
little variation in the 3dB bandwidth. But constant bandwidth is not
available for the variation of the other external resistances. This
supports our previous comments on the gain bandwidth behavior
of the toplogy-1, which we did by observing the gain equation and
there corresponding bode plots.
 84

4.4 Simulation of the CMOS design of

proposed topology – 1

The main circuit components of our proposed topology – 1

are positive second generation current conveyor (CCII+) and
current amplifier. Initially we will concentrate on the positive
second generation current conveyor (CCII+) and the current
amplifier separately.

4.4.1 Positive Second Generation Current

Conveyor (CCII+)

Following figure shows the schematic diagram of the CCII+.

Fig 4.13: Schematic diagram of the CCII+.

 85

Following table provides the aspect ratios of the MOS

devices used in the CCII+.

W/L (μm/μm)
M1, M2 100/1
M3, M4 40/1
M5, M11, M13 20/1
M6 45/1
M7 10/1
M10, M12 60/1
M8 2.7/1
M9 5/1

Table 4.1: Transistor geometries for the CCII+ schematic.

Following figure shows the input and output currents of the

CCII+ of fig-4.13.

Fig 4.14: I(I_I1) is the input current and I(R_R1) is the output
current.

From the above figure we found that the CCII+ provides zero
current for the negative half cycle of the input current and gain is
also less than zero. It may happen for improper biasing.
The schematics netlist of the circuit of the fig-4.13 is
provided in the appendix C.1.
 86

4.4.2 Current Amplifier

Following figure shows the circuit of a current amplifier

Fig 4.15: Schematic diagram of a simple open-loop bidirectional

current amplifier.

The current gain of the above current amplifier is set by the

aspect ratio of M5 (M8) and M7 (M10) –

(W / L)7 (W / L)10 ………………………………………..(4.1)

A = iout = =
iin (W / L)5 (W / L)8
 87

Following table provides the aspect ratios of the MOS

devices used in the above current amplifier.

W/L (μm/μm)
M1 20/1
M2 20/1
M3 20/1
M4 20/1
M5 20/1
M6 20/1
M7 80/1
M8 20/1
M9 20/1
M10 80/1

Table 4.2: Transistor geometries for the schematic of the current

amplifier.

Following figure shows the input and output currents of the

current amplifier of fig-4.15.

Fig 4.16: I(I1) is the input current and I(R1) is the output current.

The schematics netlist of the circuit of the fig-4.15 is

provided in the appendix C.2.
 88

4.4.3 Ultimate simulation of the CMOS design of

proposed topology – 1

Finally we combined CCII+ circuit and current amplifier

circuit to form the inverting configuration our proposed topology –
1. The circuit diagram is shown bellow.

Fig 4.17: Schematic diagram of our proposed topology – 1 in

inverting configuration.
 89

Following figure shows the input and output voltage wave

forms of the circuit of the fir-4.17.

Fig 4.18: V1(V_V6) is the input voltage and V2(R_Rload) is the

output voltage of the circuit of the fig – 4.17.

Here we can observe that output voltage wave form is

attenuated from the input voltage signal. So here the gain is less
than one. But our main objective is to obtain variable GBP. So, for
the time being we will not be bothered by the attenuated output
voltage of the circuit of the fig – 4.17.

Another thing here we can note that our output is not inverted
though it is in the inverting configuration. The reason of this
phenomenon can be realized if we compare carefully the direction
of the output current of current amplifier in both small signal
model and in the CMOS design. The comparison says that output
current of the CMOS current amplifier is in opposite direction to
that of the small signal model of the current amplifier.
 90

Now we will proceed to the AC – analysis. First we will vary

the input resistance R_Rgncntrl and observe the effects of its
variation on the 3dB bandwidth. Following three figures will show
the effects of the variation of R_Rgncntrl on the 3dB bandwidth. In
these figures we kept the value of the other external resistances
constant (R_Rload = 5kΩ and R_Rfeedback = 1kΩ).

Fig 4.19 : AC analysis with the input resistance R_Rgncntrl = 1kΩ.

Fig 4.20 : AC analysis with the input resistance R_Rgncntrl =

100Ω.
 91

Fig 4.21 : AC analysis with the input resistance R_Rgncntrl = 10Ω.

Observing the figures above we can say that voltage gain of

the circuit of the fig – 4.17 is inversely proportional to the input
resistance which is also supported by our derived gain equation.
Here we have also found an infinitely wide flat bandwidth.

Now we will observe the effect of output resistance variation

in the 3dB bandwidth. Following figure shows the AC analysis of
the circuit of the fig - 4.17 with output resistance R_Rload =
500kΩ, feedback resistance R_Rfeedback = 1kΩ and input
resistance R_Rgnctrl = 10.
 92

Fig 4.22 : AC analysis with the output resistance R_Rload =

500kΩ.
Surprisingly here we have also found an infinitely wide flat
bandwidth which is obviously not expected.

Now we will observe the effect of feedback resistance

variation in the 3dB bandwidth. Following figure shows the AC
analysis of the circuit of the fig - 4.17 with output resistance
R_Rload = 5kΩ, feedback resistance R_Rfeedback = 10Ω and
input resistance R_Rgnctrl = 10.

Fig 4.23 : AC analysis with the feedback resistance R_Rfeedback

= 10Ω.
 93

Surprisingly here we have also found an infinitely wide flat

bandwidth which is obviously not expected.

To analyze the reasons of these surprising results we will first

observe the individual AC analysis of the CCII+ circuit and the
current amplifier circuit.

Following figure shows the AC analysis of the CCII+.

Fig 4.24: AC analysis of the CCII+.

Following figure shows the AC analysis of the current

amplifier.

Fig 4.25: AC analysis of the current amplifier.

 94

Now we can say by observing the above AC analysis of the

CCII+ and current amplifier circuit that as both of their frequency
response are infinitely wide and flat, the circuit built by cascading
them will also have an infinitely wide and flat frequency response
curve. But this phenomenon is completely impractical.

Infinitely wide and flat frequency response curve of our

design may result from the fact that the MOS devices we used in
simulation were generalized and ideal in nature.

Now we will perform the simulation of our design using the

MOS devices having parameters as following -

*model specification:
*MCE 3 Micron CMOS processes parameters-process 2
*N channel type
.model nenh nmos level=2 vto=0.85 kp=30e-6 tox=470e-10 nsub=38e14
+ld=0.6e-6 uo=624 uexp=0.055 vmax=20e4 neff=9.8 delta=2.0 cj=160e-6
cjsw=430e-12 mj=0.5 mjsw=0.33 pb=0.81
*p channel typ
.model penh pmos level=2 vto=-0.85 kp=12e-6 tox=470e-10 nsub=8.7e14
+ld=0.5e-6 uo=200 uexp=0.18 vmax=12e4 neff=4.0 delta=2.0
+cj=100e-6 cjsw=180e-12 mj=0.5 mjsw=0.33 pb=0.7
.ac DEC 10000 10 10000GHz

Now we will observe the AC analysis output of our CMOS

design of the topology – 1 in inverting configuration. First we will
vary the input resistance R_Rgncntrl and observe the effects of its
variation on the 3dB bandwidth. Following two figures will show
the effects of the variation of R_Rgncntrl on the 3dB bandwidth. In
these figures we kept the value of the other external resistances
constant.
 95

Fig 4.26 : AC analysis with the input resistance R_Rgncntrl = 10Ω.

Fig 4.27 : AC analysis with the input resistance R_Rgncntrl = 1kΩ.

Observing the figures above we can say that voltage gain of

the circuit of the fig – 4.17 is inversely proportional to the input
resistance which is also supported by our derived gain equation.
But here 3dB bandwidth also varies with the variation of input
resistance, which is not expected.

Now we will observe the effect of output resistance variation

in the 3dB bandwidth. Following figures show the AC analysis of
 96

the circuit of the fig - 4.17 with other external resistances kept
constant.

Fig 4.28 : AC analysis with the feedback resistance R_Rload =

0.5kΩ.

Fig 4.29 : AC analysis with the feedback resistance R_Rload =

5kΩ.
Observing the above frequency response curves, we found
that variation of the output resistance has a little effect on the 3dB
bandwidth.
 97

Now we will observe the effect of feedback resistance

variation in the 3dB bandwidth. Following figures show the AC
analysis of the circuit of the fig - 4.17 with other external
resistances kept constant.

Fig 4.30 : AC analysis with the feedback resistance R_Rfeedback

= 10Ω.

Fig 4.31 : AC analysis with the feedback resistance R_Rfeedback

= 100Ω.
 98

Fig 4.32 : AC analysis with the feedback resistance R_Rfeedback

= 1kΩ.

Fig 4.33 : AC analysis with the feedback resistance R_Rfeedback

= 1kΩ.

Schematics netlist of the CMOS design of our topology – 1 in

inverting configuration using the latter MOS device parameters is
provided in the appendix – C.3. Observing the above frequency
response curves, we surprisingly discovered that variation of
 99

feedback resistance does not have any effect on the 3dB bandwidth.
But our gain equation and simulation of the small signal model
says that we should have obtain constant bandwidth for the
variation of input resistance to control gain.

For this surprising result we checked back for errors in the

gain equation of the topology -1 for inverting configuration. But
we could not detect any errors there. There we neglected some
parameters for their very small values. We thought that those
parameters are too small to make any noticeable effect in the gain
equation. Even though now we will reconsider the neglected
parameters again. By reconsidering the previously neglected
parameters we obtained the gain equation as following –

Gv = − AiωtRoRf
Ri{S 2CzRiA + S ( Rf + Ro + RiA + CzRiA) + [ωt ( Rf + Ro) + AiωtRo + ωt ( RiA )]}
Rz Rz
………………………(4.2)

Comparing equation – 4.2 with equation 2.1, we can see that

equation 4.2 has second order terms which equation 2.1 does not
have. Other than that equation 4.2 provides more or less same
characteristics of equation 2.1.

Now we will observe the effect of the variation of external

resistances on the gain equation 4.2. To do this again we will take
help of MATLAB.

Following are the different parameters of our proposed

topology – 1 with their values which are used as the initial input of
the MATLAB program to obtain bode plot.

Gain of the current amplifier, Ai =10

Bandwidth of the current amplifier, ωt=1000 Hz
 100

Output resistance, Ro=1k Ω

Feedback resistance Rf =100 Ω
Input resistance Ri =20Ω
Resistance at node Z, Rz =50KΩ
Capacitance at node Z, Cz =1pF
Input resistance of the current amplifier, RiA =5Ω

With the values of the parameters mentioned above we

obtained the bode plot shown below.

Fig 4.34: Bode plot for the values mentioned initially.

Now we will vary the input resistance Ri. The bode plot
below shows the variation of Ri with the other parameters remain
unchanged as defined initially. Used values of the input resistance
Ri in the bode plot are 10Ω, 100Ω, 1kΩ, 10kΩ.
 101

Fig 4.35: Bode plots showing the effect of variation of Ri.

Now we will vary the feedback resistance Rf. The bode plot
below shows the variation of Rf with the other parameters remain
unchanged as defined initially. Used values of the feedback
resistance Rf in the bode plot are 10Ω, 100Ω, 1kΩ, 10kΩ.
 102

Fig 4.36: Bode plots showing the effect of variation of Rf.

Now we will vary the output resistance Ro. The bode plot
below shows the variation of Ro with the other parameters remain
unchanged as defined initially. Used values of the output resistance
Ro in the bode plot are 10Ω, 100Ω, 1kΩ, 10kΩ.
 103

Fig 4.37: Bode plots showing the effect of variation of Ro.

Our observation on the above bode plots found that for the
variation of input resistance Ri there is no variation in the 3dB
bandwidth. But constant bandwidth is not available for the
variation the other external resistances. This supports our previous
comments on the gain bandwidth behavior of the toplogy-1, which
we did by observing the gain equation.

The MATLAB code for the above bode plots is provided in

the appendix – B.5.

So, we could not find any error in our proposed topology – 1.

 104

4.5 Conclusion

All the schematics netlists of the simulations of our CMOS

design are provided in the appendix – 3. The simulation of larger
circuits may not always converge to the end point. The simulator
of the PSpice uses some iterative numerical methods. These
iterations may not converge all the times. In this case we may
converge to the end point if we reduce the analysis time for
transient analysis or change other different parameters of the
circuit. Use of “.OPTION STEPGMIN” is also a good ploy.
 105

Chapter – 5

Future Works and Conclusion

5.1 Future Works

The research work in this thesis focused on the constant

bandwidth behavior in four possible configurations these are as
following -

*Amplifying element is current amplifier and feedback amplifier is

in voltage amplifier configuration.
*Amplifying element is voltage amplifier and feedback amplifier is
in current gain configuration.
*Amplifying elements are voltage and current amplifier and
feedback amplifier is in transimpedance configuration.

We have proved the constant bandwidth and variable gain

bandwidth product in the case of the voltage gain configuration
when the amplifying element is current amplifier by theoretical
analysis and the results achieved by software simulation. Further
exploration is required for the CMOS design of this topology to
obtain accurate result.
 106

In the gain equations derived for the last three topologies we

found a term (not equal for each topology) is present in each
topology affecting the 3dB bandwidth. The effect of these terms
can be minimized by using a special set of values of the parameters
concerning the topology. The software simulation can also be done
in the similar manner as described for the first case. That means
further exploration can be carried on for the last three topologies.

Because of some unavoidable circumstances, we were unable

to prove the gain independent bandwidth in these four cases by
hardware implementation. In future hardware implementation can
be accomplished basing on the theories we derived in this paper,
which was our desired goal.

New amplifier designs can be explored both theoretically and

experimentally, based on the variable gain bandwidth properties
and constant bandwidth when the transconductance configuration
acts as the feedback amplifier and the current and voltage amplifier
as the amplifying element.
 107

5.2 Conclusion

At first we thought that we will be able to implement the

proposed topology – 1 practically with MOS devices. But
collection of all the MOS devices having our desired aspect ratio
from the market is quite difficult. Then we tried to collect the
CCII+ IC from the market. But we also failed to collect it. We
believe that more study and research on the way we tried to obtain
constant band width will lead to the desired success. Then the
design can also be implemented in a single IC by CMOS
technology.
 108

Bibliography
[1] Sharker,Md. Ataur Rahman, “Nobel Design Approaches for
Low voltage CMOS Current Feedback and Transconductance
Feedback Amplifiers, August 2004.

[2] R. F. Coughlin and F. F. Driscoll, “Operational Amplifiers and

Linear Integrated Circuits”, 6th ed. Prentice Hall, New Jersey,
2001.

[3] D. A. Pucknell and K. Eshraghian, Basic VLSI Design, 3rd ed.

Prentice Hall,
New Delhi, 1999.

[4] J. Millman and C. C. Halkias, “Integrated Electronics: Analog

and Digital Circuits and Systems”, fourth reprint ed. McGraw Hill,
India, 1992.

[5] Robert Boylestad and Louis Nashelsky, “Electronic Devices

and Circuit Theory”, sixth edition, Prentice Hall of India,1997.

[6] Adel S. Sedra, Kenneth C. Smith,” Microelectronic Circuits”,

fifth edition,oxford university press, 2005-2006.

[7] Paul R. Gray, Paul J. Hurst, Stephen H.Lewis, Robert G. Meyer,

“Analysis and Design of Analog Integrated circuits”, fourth edition.
Jhon Wiely & sons Ltd

[8] B. J. Maundy, A. R. Sarkar, and S. J. Gift, “A new design

topology for low voltage CMOS current feedback amplifiers,”
IEEE Transactions on Circuits and
Systems II: Analog and Digital Signal Processing, (in press), 2004.
 109

[9] B. J. Maundy, A. R. Sarkar, and S. J. Gift, “A new topology for

the transconductance feedback amplifiers,” IEEE Transactions on
Circuits and
Systems II: Analog and Digital Signal Processing.

[10] Current Feedback Amplifier Theory and Applications,

Application Note 9420.1, Intersil Corporation, April 1995.

[11] S. Franco, Design with Operational Amplifiers and Analog

Integrated Circuits, 3rd ed. McGraw Hill, Boston, 2002.

[12] B. Wilson, “Trends in current conveyor and current-mode

amplifier design,”
International Journal of Electronics, vol. 73, pp. 573–583,
September 1992.

[13] B. Wilson, “Generalized analysis of gain-bandwidth

independence in feedback amplifiers,” International Journal of
Electrical Engineering Education, vol. 39, pp. 20–30, January 2002.

[14] —, “Bandwidth behavior of negative-feedback amplifiers with

dominant pole
compensation,” International Journal of Electrical Engineering
Education,
vol. 36, pp. 91–101, April 1999.

[15] B. Wilson, “High-performance current conveyor

implementation,” Electronics
Letters, vol. 20, pp. 990–991, 1984.

[16] Giuseppe Palmisano, Gaetano Palumbo, Salvatore

Pennisi,”Cmos Current Amplifiers”.
 110

Appendix A
Formula Derivations

A.1 Formulas of proposed topology – 1

Inverting Configuration:

Fig 3.2: proposed topology – 1 small signal model in inverting

configuration.

Input current to current conveyor,

iin = Vin ; ……………………………………………………..(1)

Ri
Feedback current, if = Vo ;…………………...…………………………………...(2)
Rf
 111

∴Using KCL at node –X ,

iin + if + ix = 0
⇒ ix = −(iin + if )
⇒ ix = −(Vin + Vo )
Ri Rf
…………………………………………(3)

Current input to the current amplifier block,

iiA = ((1/ SCz)|| Rz) × ix
RiA + ((1/ SCz)|| Rz)
Rz
= 1+ SCzRz × ix
RiA + Rz
1+ SCzRz
= Rz × ix
( RiA + Rz) + SCzRzRiA
= 1 × ix
R
(1+ ) + SCzRiA
iA
Rz
……………….....………………………………….(4)

As RiA is input resistance of the current amplifier, we can assume

that RiA << Rz. So we can neglect (RiA / Rz) from equation – 4.

∴From equation – 2 :

iiA ≈ 1 × ix ………………………………(5)
1+ SCzRiA
The output resistance (Rio) of the current amplifier is very high, so
considered it as an open circuit.

∴Output current,
iout ≈ Ai(S )iiA
………………………………………………………(6)
= Ai(S )ix
1+ SCzRiA
 112

And,
iout = io + if
⇒ io = iout − if
………………………………………………………………….(7)

Now voltage across the resistance Ro,

Vo = Ro × io
= Ro(iout − if )
[From equation - 7]

= Ro × ( Ai(S )ix − if ) [ From equation - 6]

1+ SCzRiA
Ai(S )
= Ro ×[ {−(Vin + Vo )} − Vo ]
1+ SCzRiA Ri Rf Rf
[From equation–3 & 2]
Ai(S ) Ro Ai(S ) Ro
⇒ Vo ×[1+ + Ro ] = −Vin ×
Rf (1+ SCzRiA) Rf Ri(1+ SCzRiA)

Ai(S ) Ro
⇒ Vo = − Ri(1+ SCzRiA)
Vin [ Rf (1+ SCzRiA) + Ai(S ) Ro + Ro(1+ SCzRiA) ]
Rf (1+ SCzRiA)

Ai(S ) RoRf
=−
Ri{Rf (1+ SCzRiA) + Ai(S ) Ro + Ro(1+ SCzRiA)}

Ai(S ) RoRf
=− ………………(8)
Ri{( Rf + Ro) + Ai(S ) Ro + ( Rf + Ro)(1+ SCzRiA)}
 113

Since Cz is very small and RiA is also very small, we can

consider SCzRiA is close to zero and negligible. Then equation –
(8) becomes –
Vo ≈ − Ai(S ) RoRf ……………………………….(9)
Vin Ri{( Rf + Ro) + Ai(S ) Ro}

The current amplifier gain Ai(S) is assumed to be single pole roll-

off model, given by –
Ai(S ) = Aiωt …………………………………………………...….(10)
S + ωt

∴From equation 9 & 10 we can write –

Aiωt × RoRf
Gv = Vo = − S + ωt
Vin Ri{( Rf + Ro) + Aiωt × Ro}
S + ωt
=− AiωtRoRf
Ri{(S + ωt )( Rf + Ro) + AiωtRo}
…………….…………………….(11)

We know that at the 3dB point the real part and the imaginary part
of the gain equation will become equal.

∴From equation (11) for 3dB band width –

Real part = Imaginary part

⇒ ω ( Rf + Ro) = ωt ( Rf + Ro) + AiωtRo [Putting S=jω]

⇒ f ( Rf + Ro) = ft ( Rf + Ro) + Ai ⋅ ft ⋅ Ro

⇒ f = ft + Ai ⋅ ft ⋅ Ro
( Rf + Ro)
∴For the inverting configuration the 3dB bandwidth is -
f = ft{1+ Ai ⋅ Ro } ………………………………………………………..(12)
Rf + Ro
 114

A.2 Formulas of proposed topology – 2

Inverting Configuration:

Fig 3.5: proposed topology – 2 small signal model in inverting

configuration.

Input current to current conveyor is equal to iin .

Feedback current, if = Vo ;………………………………………………...(13)

Rf
∴Using KCL at node –X ,

iin + if + ix = 0
⇒ ix = −(iin + if )
⇒ ix = −(iin + Vo )
Rf
…………………………………………(14)

Here we assumed ideal cases for the voltage amplifier

that is - we assumed input resistance (RiA) of the voltage
amplifier tends to infinity and output resistance (RoA) tends to
zero. So we considered that a negligible amount of current
passes through input resistance (RiA) and the voltage drop
across the output resistance (R oA ) is also negligible.
 115

VA is the input voltage of the voltage amplifier. It is the

voltage across ((1/ SCz ) || Rz ) .

As the current flowing through input resistance (RiA) is

negligible –

VA = ((1/ SCz)|| Rz) × ix

= −(iin + Vo ) × Rz ………………………………………..(15)
Rf 1+ SCzRz
Let the gain of the voltage amplifier is Ao(S).

As voltage drop across the output resistance (RoA) is negligible,

output voltage (VAo) can be given as –

VAo = Ao(S )VA

RzAo(S )
= −(iin + Vo ) × ………………………………………….(16)
Rf 1+ SCzRz

We assumed ideal cases for the transconductance block that

is - we assumed input resistance (Rig) and output resistance (Rog) of
the transconductance block tends to infinity. So we considered that
a negligible amount of current passes through both the input
resistance (RiA) and the output resistance (RoA). Let the gain of the
transconductance block is gm(S).

∴Output current (iog) can be given as –

igo = gm(S )VAo

RzAo(S ) gm(S )
= −(iin + Vo ) × ………………………………………………..(17)
Rf 1+ SCzRz
 116

As the current flowing through the output resistance (Rog) is

negligible, we can write KCL at the output node as following

igo = io + if
⇒ io = igo − if
RzAo(S ) gm(S ) Vo Ro
⇒ io = −(iin + Vo Ro ) × ( )− [From equation (17) & (13)]
Ro Rf 1+ SCzRz Ro Rf
RzAo(S ) gm(S )
⇒ io = −(iin + io × Ro ) × ( ) − io × Ro [∵io = Vo ]
Rf 1+ SCzRz Rf Ro
RzAo(S ) gm(S ) Ro RzAo(S ) gm(S )
⇒ io{1+ Ro × + } = −iin × ( )
Rf 1+ SCzRz Rf 1+ SCzRz

RzAo(S ) gm(S )
i
⇒ =−
o 1+ SCzRz
iin Rf (1+ SCzRz) + RoRzAo(S ) gm(S ) + Ro(1+ SCzRz)
Rf (1+ SCzRz)

RfRzAo(S ) gm(S )
⇒ io = − ……………(18)
iin Rf (1+ SCzRz) + RoRzAo(S ) gm(S ) + Ro(1+ SCzRz)

The voltage amplifier gain Ao(S) is assumed to be single pole

roll-off model, given by –

Ao(S ) = Aoωt ………………………………………………….….(19)

S + ωt

The bandwidth of the voltage amplifier has to be taken much

larger than the transconductance block. For the transconductance
block we can take gm(S) = gm.
 117

∴From equation (18) the current gain –

Gi = io = − RfRzAoωtgm
iin Rf (1+ SCzRz)(S + ωt ) + RoRzAoωtgm + Ro(1+ SCzRz)(S + ωt )

=− RfRzAoωtgm
RoRzAoωtgm + ωt ( Rf + Ro) + S (1+ ωtCzRz)( Rf + Ro) + S 2CzRz ( Rf + Ro)

………………………...(20)

Putting S=jω in the equation (20) –

Gi = − RfRzAoωtgm
RoRzAoωtgm + ωt ( Rf + Ro) + jω (1+ ωtCzRz)( Rf + Ro) − ω 2CzRz( Rf + Ro)

..............................……(21)

We know that at the 3dB point the real part and the
imaginary part of the gain equation will become equal.

∴From equation (21) for 3dB band width –

Real part = Imaginary part

⇒ ω (1+ ωtCzRz)( Rf + Ro) = −ω 2CzRz( Rf + Ro) + ωt ( Rf + Ro) + RoRzAoωtgm

⇒ ω 2CzRz( Rf + Ro) + ω (1+ ωtCzRz)( Rf + Ro) −{ωt ( Rf + Ro) + RoRzAoωtgm} = 0

⇒ ω 2CzRz + ω (1+ ωtCzRz) −{ωt + RoRzAoωtgm} = 0

( Rf + Ro)
−(1+ ωtCzRz) ± (1+ ωtCzRz)2 + 4CzRz ×{ωt + RoRzAoωtgm}
( Rf + Ro)
⇒ω =
2CzRz
 118

−(1+ 2π ftCzRz) ± (1+ 2π ftCzRz)2 + 8π ftCzRz ×{1+ RoRzAogm}

( Rf + Ro)
⇒f=
4π CzRz
………………………………(22)

A.3 Formulas of proposed topology – 3

Inverting Configuration:

Fig 3.8: proposed topology – 3 small signal model in inverting

configuration.

Input current to current conveyor is equal to iin .

Feedback current,

if = Vo ; ……………………………………………...(23)
Rf
 119

∴Using KCL at node –X ,

iin + if + ix = 0
⇒ ix = −(iin + if )
⇒ ix = −(iin + Vo )
Rf
…………………………..……………………(24)

The internal capacitance at the Z node of the CCII+ is

Cz and the resistance at the Z node is Rz.

Here we assumed ideal cases for the voltage amplifier

that is - we assumed input resistance (RiA) of the voltage
amplifier tends to infinity and output resistance (RoA) tends to
zero. So we considered that a negligible amount of current
passes through input resistance (RiA) and the voltage drop
across the output resistance (RoA) is also negligible.

VA is the input voltage of the voltage amplifier. It is the

voltage across ((1/ SCz )|| Rz) .

As the current flowing through input resistance (RiA) is

negligible –
VA = ((1/ SCz)|| Rz) × ix

= −(iin + Vo ) × Rz …………………………………………..(25)
Rf 1+ SCzRz

Let the gain of the voltage amplifier is Ao(S).

As voltage drop across the output resistance (RoA) is negligible,
output voltage (Vo) can be given as –

Vo = Ao(S )VA
RzAo(S )
= −(iin + Vo ) × ………………………………………….(26)
Rf 1+ SCzRz
 120

RzAo(S ) RzAo(S )
⇒ Vo{1+ } = −iin ×
Rf (1+ SCzRz) 1+ SCzRz

Rf (1+ SCzRz) + RzAo(S ) RzAo(S )

⇒ Vo{ } = −iin ×
Rf (1+ SCzRz) 1+ SCzRz

RfRzAo(S )
⇒ Vo = − ………………………………….(27)
iin Rf (1+ SCzRz) + RzAo(S )

The voltage amplifier gain Ao(S) is assumed to be single pole roll-

off model, given by –

Ao(S ) = Aoωt ……………………………………………….….(28)

S + ωt

From equation (27) gain of the transimpedance amplifier –

⇒ Rm = V o = − RfRzAoωt
iin Rf (1+ SCzRz)(S + ωt ) + RzAoωt

=− RfRzAoωt ………………….(29)
Rf ωt + RzAoωt + S 2 RfCzRz + SRf (1+ ωtCzRz)

Putting S=jω in the equation (29) –

Rm = Vo = − RfRzAoωt ……(30)
iin Rf ωt + RzAoωt − ω RfCzRz + jω Rf (1+ ωtCzRz)
2

We know that at the 3dB point the real part and the imaginary part
of the gain equation will become equal.

∴From equation (30) for 3dB band width –

 121

Real part = Imaginary part

⇒ ω Rf (1+ ωtCzRz) = Rf ωt + RzAoωt − ω 2 RfCzRz

⇒ ω 2 RfCzRz + ω Rf (1+ ωtCzRz) −{ωt ( Rf + RzAo)} = 0
− Rf (1+ ωtCzRz) ± Rf 2 (1+ ωtCzRz)2 + 4CzRzRf ωt ( Rf + RzAo)
⇒ω =
2CzRzRf
− Rf (1+ 2π ftCzRz) ± Rf 2 (1+ 2π ftCzRz)2 + 8π ftCzRzRf ( Rf + RzAo)
⇒f=
4π CzRzRf
………..(31)

A.4 Formulas of proposed topology – 4

Inverting Configuration:

Fig 3.11: proposed topology – 4 small signal model in inverting

configuration.

Input current to current conveyor is equal to iin .

Feedback current, if = Vo ;……………………………………………………...(32)
Rf
 122

∴Using KCL at node –X ,

iin + if + ix = 0
⇒ ix = −(iin + if )
⇒ ix = −(iin + Vo )
Rf
………………………………...………………(33)

Here we assumed ideal cases for the current amplifier

that is - we assumed input resistance (RiA) of the current
amplifier is small and output resistance (RoA) tends to infinity.
So we considered that a negligible amount of current passes
through output resistance (RoA).

Current input to the current amplifier block,

iiA = ((1/ SCz)|| Rz) × ix

RiA + ((1/ SCz)|| Rz)
Rz
= 1+ SCzRz × ix
RiA + Rz
1+ SCzRz
= Rz × ix
( RiA + Rz) + SCzRzRiA
= 1 × ix
R
(1+ ) + SCzRiA
iA
Rz
……………………………………..……….(34)

As RiA is input resistance of the current amplifier, we can assume

that RiA << Rz. So we can neglect (RiA / Rz) from equation -34.

∴From equation – 34 :
iiA ≈ 1 × ix ………………………………………………(35)
1+ SCzRiA
 123

The output resistance (RoA) of the current amplifier is very high, so

considered it as an open circuit.
∴Output current,
ioA ≈ Ai(S )iiA
= Ai(S )ix
1+ SCzRiA
………………..……………………………………(36)

We assumed ideal cases for the transimpedance block that is

- we assumed input resistance (Rig) and output resistance (Rog) of
the transimpedance block tends to Zero. So we considered that
voltage drop across the output resistance (RoA) is negligible. Let
the gain of the transimpedance block is Rm(S).

∴The output voltage –

Vo = Rm(S ) × ioA
Ai(S )ix
= Rm(S ) ×
1+ SCzRiA

Rm(S ) Ai(S )
= −(iin + Vo ) ×
Rf 1+ SCzRiA
Rm(S ) Ai(S ) Rm(S ) Ai(S )
⇒ Vo{1+ } = −iin ×
Rf (1+ SCzRiA) 1+ SCzRiA

Rf (1+ SCzRiA) + Rm(S ) Ai(S ) Rm(S ) Ai(S )

⇒ Vo{ } = −iin ×
Rf (1+ SCzRiA) 1+ SCzRiA

RfRm(S ) Ai(S )
⇒ Vo = − ……………………………………(37)
iin Rf (1+ SCzRiA) + Rm(S ) Ai(S )

The current amplifier gain Ai(S) is assumed to be single pole roll-

off model, given by –

Ai(S ) = Aiωt ………………………………………………….….(38)

S + ωt
 124

The bandwidth of the current amplifier has to be taken much

larger than the transimpedance block. For the transimpedance
block we can take Rm(S) = Rm.

∴From equation (37) the transimpedance gain –

Rm = Vo = − RfRmAiωt
iin Rf (1+ SCzRiA)(S + ωt ) + RmAiωt

=− RfRmAiωt ………………………(39)
Rf ωt + RmAiωt + S CzRiARf + SRf (1+ CzRiAωt )
2

Putting S=jω in the equation (39) –

Rm = − RfRmAiωt ………………..(40)
Rf ωt + RmAiωt − ω 2CzRiARf + jω Rf (1+ CzRiAωt )

We know that at the 3dB point the real part and the
imaginary part of the gain equation will become equal.

∴From equation (40) for 3dB band width –

Real part = Imaginary part

⇒ ω Rf (1+ CzRiAωt ) = Rf ωt + RmAiωt − ω 2CzRiARf

⇒ ω 2CzRiARf + ω Rf (1+ CzRiAωt ) − ωt ( Rf + RmAi) = 0

− Rf (1+ CzRiAωt ) ± Rf 2 (1+ CzRiAωt )2 + 4CzRiARf ωt ( Rf + RmAi)

⇒ω =
2CzRiARf
− Rf (1+ 2π ftCzRiA) ± Rf 2 (1+ 2π ftCzRiA)2 + 8π ftCzRiARf ( Rf + RmAi)
⇒f=
4π CzRiARf
…………(41)
 125

Appendix B

MATLAB Codes Used

B.1 Codes used for topology – 1.

clear all
Prm=input('Enter Ai, Wt, Ro, Rf, Ri : ');
[M0,N0,N1]=params_config1(Prm);
g=tf([M0],[N1 N0]);
bode(g)
grid on
figure
rstr=Prm(5);
for x=1:1:4
Prm(5)=10^(x);
[M0_,N0_,N1_]=params_config1(Prm);
g=tf([M0_],[N1_ N0_]);
bode(g)
title('Varing Ri');
grid on
hold on
end
legend('1st','2nd','3rd','4th',4);
Prm(5)=rstr;
rstr1=Prm(4);
figure
for m=1:1:4
Prm(4)=10^(m);
[M0_1,N0_1,N1_1]=params_config1(Prm);
g=tf([M0_1],[N1_1 N0_1]);
bode(g)
title('Varing Rf');
grid on
hold on
end
legend('1st','2nd','3rd','4th',4);
 126

Prm(4)=rstr1;
figure
for n=1:1:4
Prm(3)=10^(n);
[M0_2,N0_2,N1_2]=params_config1(Prm);
g=tf([M0_2],[N1_2 N0_2]);
bode(g)
title('Varing Ro');
grid on
hold on
end
legend('1st','2nd','3rd','4th',4);

function[M0,N0,N1]=params_config1(Prm)
M0=-1*(Prm(1)*Prm(2)*Prm(3)*Prm(4))/Prm(5);
N0=Prm(2)*(Prm(4)+Prm(3)+Prm(1)*Prm(3));
N1=Prm(4)+Prm(3);

B.2 Codes used for topology – 2.

clear all
Prm=input('Enter Ao, Wt, Gm, Cz, Rz, Ro, Rf : ');
[M0,N0,N1,N2]=params_config2(Prm);
g=tf([M0],[N2 N1 N0]);
bode(g)
grid on
figure
rstr1=Prm(7);
for m=1:1:4
Prm(7)=10^(m);
[M0_1,N0_1,N1_1,N2_1]=params_config2(Prm);
g=tf([M0_1],[N2_1 N1_1 N0_1]);
bode(g)
title('Varing Rf');
grid on
hold on
end
legend('1st','2nd','3rd','4th',4);
Prm(7)=rstr1;
 127

figure
for n=1:1:4
Prm(6)=10^(n);
[M0_2,N0_2,N1_2,N2_2]=params_config2(Prm);
g=tf([M0_2],[N2_2 N1_2 N0_2]);
bode(g)
title('Varing Ro');
grid on
hold on
end
legend('1st','2nd','3rd','4th',4);

function[M0,N0,N1,N2]=params_config2(Prm)
M0=-1*Prm(7)*Prm(5)*Prm(1)*Prm(2)*Prm(3);
N0=Prm(6)*Prm(5)*Prm(1)*Prm(2)*Prm(3)+Prm(2)*(Prm(7)+Prm(6));
N1=(1+Prm(4)*Prm(5)*Prm(2))*(Prm(7)+Prm(6));
N2=Prm(4)*Prm(5)*(Prm(7)+Prm(6));

B.3 Codes used for topology – 3.

clear all
Prm=input('Enter Ao, Wt, Cz, Rz, Rf : ');
[M0,N0,N1,N2]=params_config3(Prm);
g=tf([M0],[N2 N1 N0]);
bode(g)
grid on
figure
for m=1:1:4
Prm(5)=10^(m);
[M0_1,N0_1,N1_1,N2_1]=params_config3(Prm);
g=tf([M0_1],[N2_1 N1_1 N0_1]);
bode(g)
title('Varing Rf');
grid on
hold on
end
legend('1st','2nd','3rd','4th',4);
 128

function[M0,N0,N1,N2]=params_config3(Prm)
M0=-1*Prm(5)*Prm(4)*Prm(1)*Prm(2);
N0=Prm(2)*(Prm(5)+Prm(4)*Prm(1));
N1=Prm(5)*(1+Prm(3)*Prm(4)*Prm(2));
N2=Prm(3)*Prm(4)*Prm(5);

B.4 Codes used for topology – 4.

clear all
Prm=input('Enter Ai, Wt, ,Rm, Cz, Ria, Rf : ');
[M0,N0,N1,N2]=params_config4(Prm);
g=tf([M0],[N2 N1 N0]);
bode(g)
grid on
figure
for m=1:1:4
Prm(6)=10^(m);
[M0_1,N0_1,N1_1,N2_1]=params_config4(Prm);
g=tf([M0_1],[N2_1 N1_1 N0_1]);
bode(g)
title('Varing Rf');
grid on
hold on
end
legend('1st','2nd','3rd','4th',4);

function[M0,N0,N1,N2]=params_config4(Prm)
M0=-1*Prm(6)*Prm(3)*Prm(1)*Prm(2);
N0=Prm(2)*(Prm(6)+Prm(3)*Prm(1));
N1=Prm(6)*(1+Prm(4)*Prm(5)*Prm(2));
N2=Prm(4)*Prm(5)*Prm(6);

B.5 Codes used for topology – 1 (considering

neglected terms).
clear all
Prm=input('Enter Ai, Wt, Ro, Rf, Ri ,Ria ,Rz ,Cz: ');
[M0,N0,N1,N2]=params_config1(Prm);
g=tf([M0],[N2 N1 N0]);
bode(g)
 129

grid on
figure
rstr=Prm(5);
for x=1:1:4
Prm(5)=10^(x);
[M0_,N0_,N1_,N2_]=params_config1(Prm);
g=tf([M0_],[N2_ N1_ N0_]);
bode(g)
title('Varing Ri');
grid on
hold on
end
legend('1st','2nd','3rd','4th',4);
Prm(5)=rstr;
rstr1=Prm(4);
figure
for m=1:1:4
Prm(4)=10^(m);
[M0_1,N0_1,N1_1,N2_1]=params_config1(Prm);
g=tf([M0_1],[N2_1 N1_1 N0_1]);
bode(g)
title('Varing Rf');
grid on
hold on
end
legend('1st','2nd','3rd','4th',4);
Prm(4)=rstr1;
figure
for n=1:1:4
Prm(3)=10^(n);
[M0_2,N0_2,N1_2,N2_2]=params_config1(Prm);
g=tf([M0_2],[N2_2 N1_2 N0_2]);
bode(g)
title('Varing Ro');
grid on
hold on
end
legend('1st','2nd','3rd','4th',4);

function[M0,N0,N1,N2]=params_config1(Prm)
M0=-1*(Prm(1)*Prm(2)*Prm(3)*Prm(4))/Prm(5);
N0=Prm(2)*(Prm(4)+Prm(3))+Prm(1)*Prm(2)*Prm(3)+Prm(2)*(Prm(6)/Prm(7));
N1=Prm(4)+Prm(3)+(Prm(6)/Prm(7))+Prm(8)*Prm(6);
N2=Prm(8)*Prm(6);
 130

Appendix C

Schematic Netlists

C.1 Schematic netlist of the positive second

generation current conveyor (CCII+)
************************************************************************
V_V1 $N_0001 0 700mV
R_R1 0 $N_0002 1k
M_M1 $N_0004 chk1 $N_0003 $N_0003 MbreakN
+ L=1um
+ W=100um
M_M2 $N_0005 0 $N_0003 $N_0003 MbreakN
+ L=1um
+ W=100um
M_M3 $N_0004 $N_0004 $N_0006 $N_0006 MbreakP
+ L=1um
+ W=40um
M_M4 $N_0005 $N_0004 $N_0006 $N_0006 MbreakP
+ L=1um
+ W=40um
M_M5 $N_0003 $N_0001 0 0 MbreakN
+ L=1um
+ W=20um
M_M6 $N_0006 $N_0005 $N_0007 $N_0007 MbreakN
+ L=1um
+ W=45um
M_M7 $N_0007 $N_0001 0 0 MbreakN
+ L=1um
+ W=10um
M_M8 $N_0008 $N_0006 chk1 chk1 MbreakN
+ L=1um
+ W=2.7um
 131

M_M9 $N_0008 0 chk1 chk1 MbreakP

+ L=1um
+ W=5um
M_M10 chk1 $N_0005 $N_0006 $N_0006 MbreakP
+ L=1um
+ W=60um
M_M11 chk1 $N_0007 0 0 MbreakN
+ L=1um
+ W=20um
M_M12 $N_0002 $N_0005 $N_0006 $N_0006 MbreakP
+ L=1um
+ W=60um
M_M13 $N_0002 $N_0007 0 0 MbreakN
+ L=1um
+ W=20um
C_C1 $N_0005 $N_0008 0.5pF
I_I1 chk1 0 DC 0A AC 0.7071mA
+SIN 0A 1mA 5kHz 0 0 0
V_V3 $N_0006 0 2V
.lib
.ac LIN 10000 10HZ 20GHZ
.tran 100ns 1ms
.OP
.probe
.END

C.2 Schematic netlist of the current amplifier

************************************************************************
M_M8 $N_0001 $N_0001 +V +V MbreakP
+ L=1u
+ W=20u
M_M10 $N_0002 $N_0001 +V +V MbreakP
+ L=1u
+ W=80u
M_M9 $N_0003 $N_0001 +V +V MbreakP
+ L=1u
 132

+ W=20u
M_M2 $N_0001 $N_0003 $N_0004 $N_0004 MbreakN
+ L=1u
+ W=20u
M_M1 $N_0006 $N_0005 $N_0004 $N_0004 MbreakP
+ L=1u
+ W=20u
M_M6 $N_0005 $N_0006 -V -V MbreakN
+ L=1u
+ W=20u
M_M5 $N_0006 $N_0006 -V -V MbreakN
+ L=1u
+ W=20u
M_M7 $N_0002 $N_0006 -V -V MbreakN
+ L=1u
+ W=80u
I_I1 $N_0004 0 DC 0mA AC 0.7071mA
+SIN 0V 1mA 100Hz 0 0 0
M_M3 $N_0005 $N_0005 0 0 MbreakP
+ L=1u
+ W=20u
M_M4 $N_0003 $N_0003 0 0 MbreakN
+ L=1u
+ W=20u
R_R1 0 $N_0002 5k
V_V2 0 -V 5V
V_V1 +V 0 5V
.lib
.tran 1us 0.1s
.probe
.OP
.END
 133

C.3 Schematic netlist of the CMOS design of

topology – 1 for inverting configuration with the
MOS device parameters used latter.
************************************************************************
*spice option:
.width out=80
.OPTION STEPGMIN
C_C2 $N_0001 $N_0002 0.5pF
M_M16 $N_0003 $N_0003 +V +V penh
+ L=1um
+ W=40um
M_M17 $N_0001 $N_0003 +V +V penh
+ L=1um
+ W=40um
M_M19 +V $N_0001 $N_0004 $N_0004 nenh
+ L=1um
+ W=45um
V_V4 $N_0005 0 700mV
M_M18 $N_0006 $N_0005 0 0 nenh
+ L=1um
+ W=20um
M_M20 $N_0004 $N_0005 0 0 nenh
+ L=1um
+ W=10um
M_M15 $N_0001 0 $N_0006 $N_0006 nenh
+ L=1um
+ W=100um
M_M26 $N_0007 $N_0004 0 0 nenh
+ L=1um
+ W=20um
M_M25 $N_0007 $N_0001 +V +V penh
+ L=1um
+ W=60um
M_M34 $N_0008 $N_0008 0 0 nenh
+ L=1u
+ W=20u
M_M35 $N_0009 $N_0009 0 0 penh
+ L=1u
+ W=20u
M_M30 $N_0010 $N_0009 $N_0007 $N_0007 penh
+ L=1u
+ W=20u
M_M27 $N_0011 $N_0008 $N_0007 $N_0007 nenh
 134

+ L=1u
+ W=20u
M_M29 $N_0011 $N_0011 +V +V penh
+ L=1u
+ W=20u
V_V7 +V 0 5V
M_M24 chk1 $N_0004 0 0 nenh
+ L=1um
+ W=20um
M_M22 $N_0002 0 chk1 chk1 penh
+ L=1um
+ W=5um
M_M21 $N_0002 +V chk1 chk1 nenh
+ L=1um
+ W=2.7um
M_M23 chk1 $N_0001 +V +V penh
+ L=1um
+ W=60um
M_M14 $N_0003 chk1 $N_0006 $N_0006 nenh
+ L=1um
+ W=100um
M_M28 $N_0008 $N_0011 +V +V penh
+ L=1u
+ W=20u
M_M36 $N_0012 $N_0011 +V +V penh
+ L=1u
+ W=80u
R_Rload 0 $N_0012 0.5k
R_Rfeedback chk1 $N_0012 1k
R_Rgncntrl $N_0013 chk1 100
V_V6 $N_0013 0 DC 0v AC 0.7071v
+SIN 0v 1v 5kHz 0 0 0
M_M31 $N_0009 $N_0010 0 0 nenh
+ L=1u
+ W=20u
M_M33 $N_0012 $N_0010 0 0 nenh
+ L=1u
+ W=80u
M_M32 $N_0010 $N_0010 0 0 nenh
+ L=1u
+ W=20u
*model specification:
*MCE 3 Micron CMOS processes parameters-process 2
*N channel typ
.model nenh nmos level=2 vto=0.85 kp=30e-6 tox=470e-10 nsub=38e14
 135

+ld=0.6e-6 uo=624 uexp=0.055 vmax=20e4 neff=9.8 delta=2.0 cj=160e-6 cjsw=430e-12

mj=0.5 mjsw=0.33 pb=0.81
*p channel typ
.model penh pmos level=2 vto=-0.85 kp=12e-6 tox=470e-10 nsub=8.7e14
+ld=0.5e-6 uo=200 uexp=0.18 vmax=12e4 neff=4.0 delta=2.0
+cj=100e-6 cjsw=180e-12 mj=0.5 mjsw=0.33 pb=0.7
.ac DEC 10000 10 10000GHz
.OP
.probe
.END